Skip to main content

Full text of "On the photochemistry of the retina and on visual purple"

See other formats


Date Due 










APT- 


-4 1963 M 


-^OV 


1 ;^ooi) 










fDS- 


l-^i§67^ 


F 




MT^'F") 


"■ 'lJ^Z-V- 






^h 


^Jl i '■"'^ 






\n\£i 


■ ■ s " ^ - , tj. 


5!H^ 




1 fcl.iiJr^ 


*• -^ ..vw 














w 


fiTkMllllliiirilMMI 








rfitimv 






niic 


llUIJiy 






































PRINTeO IN 


U. S. A. 


^ 


NO. 23233 




QP 479.K95""' ""'™'*'*y '■"'"'Tr 
°"iiiiifiiilfitt?ite,i?.'!l3f,.,P' ""e retina and 




3 1924 024 829 750 




Cornell University 
Library 



The original of tliis book is in 
tlie Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 



http://www.archive.org/details/cu31924024829750 




QP 
-47 3 

k35 




ON 



THE PHOTOCHEMISTRY OF 



THE RETINA 



VISUAL PURPLE. 



ON 

THE PHOTOCHEMISTEY OF 
THE EETINA 



AND ON 



VISUAL PURPLE, 



TRANSLATED FROM THE GERMAN 



DE W. KtTHNE, 

FEOrESSOB OF PHYSIOLOGY IN THE UNIVERSITY OF HEIDELBEBO. 



EDITED, WITH NOTES, 



MICHAEL FOSTEE, M.D. F.RS. 

FELLOW . AND PR^LEOIOK, TRINITY COLLEGE, CAMBRIDGE. 



Honiron: 
MACMILLAN AND CO. 

1878 

[All Eights reserved.] 



(9P 



sn 



PBINTED BY 0. J. CLAY, M.A., 

AT THE UNIVBESITY PEBSS. 



PKEFACE. 



It is now more than twelvemonths since I undertook to pub- 
lish an English translation of Prof. Kiihne's treatises on The 
Photochemistry of the Retina and on Visual Purple. Circum- 
stances 'over which I have had no control' have, until now, 
prevented my intention from being accomplished. Meanwhile 
further investigations have largely changed the aspects and 
the prospects of the subject. "When Boll first announced his 
discovery of the sensitive purple colour of the retina he ex- 
pressed the very natural conviction that it was directly con- 
nected with the act of vision; and, as will be seen from the 
following pages, the earlier results obtained by Kiihne promised 
at first to aiford a facile solution of many visual problems. 
As, however, thanks to the energetic labours of this latter 
observer' and his talented assistant Dr A. Ewald, rapid pro- 

^ As the reader is possiWy aware, Kiilme has in some c[iiarters been re- 
proached for having made use of Boll's discovery. To BoU must undoubtedly 
be given the credit of the discovery that the retinas of many animals possess a 
sensitive purple colour (for the mention of it by previous observers (see p. 15) 
does not amount to a discovery), and he therefore will rightly share in all the 
fame belonging to the subsequent developement of that discovery. At the same 
time the study of Boll's writings can leave no doubt on the mind of the candid 
reader that at first Boll did not realize that the visual purple undergoes 



vi PREFACE. 

gress was made in our knowledge of the matter, it became 
step by step more and more evident that the mountain top 
was far more distant than had at first been imagined. From 
the first Kiihne seems to have felt, and he very early (see 
p. 18) stated his conviction that the visual purple was not 'the 
visual substance,' a knowledge of the changes of which would 
at once enable us to bridge over the gap between the waves 
of the luminous ether and the waves of the visual nervous im- 
pulses, but rather a type, shewing us, in a dim way, the 
manner in which the problems of the existence and behaviour 
of such a visual substance or such visual substances might 
be attacked; and, in the minds of the thoughtless, he may 
seem, in demonstrating (see Appendix, Note E) the possi- 
bility of vision and even colour vision without visual purple, 
to have destroyed, with his own hands, the value of his own 
handiwork. 

More thoughtful consideration, however, will lead us to 
see reason to hope, rather than to despair, in such an apparent 
failure. The discovery of the visual purple, in the form in which 
it was first laid hold of by the general public, seemed, as peo- 
ple say, 'too good to be true ;' and it is no little consolation to 
know that the first premature delusions have been corrected with 
the least possible delay. Gn the other hand there, remains the 
important fact that we can now point to positive photochemical 

changes through the action of light after the death of the animal; he attri- 
buted the bleaching after death to post mortem decomposition. It was Kuhne 
who was the first to observe that light bleaches the visual purple after death. 
This discovery made quite independently of BoU and in contradiction to Boll's 
statements is ample justification for Kiilme's continuing his investigations 
with a view to determining the bearings of what he 'himself had found 
out.' When the following pages have been read, it will be at once seen that 
the bleaching of the retina by light after death is the key to the whole matter. 
In the second edition of my Text-Booh of Physiology (p. 414, 1. 19 from bottom 
small print, "which after a few seconds exposure to Kght," &c.) the desire to 
condense the greatest amount of information into as few lines as possible, has 
led me to use a form of expression which makes me seem to attribute to Boll 
what really belongs to Kiihne : I gladly take this opportunity of correcting_this 
inadvertence. 



PREFACE. vil 

actions taking place in the retina : the first stone of a true 
theory of vision (i. e. of the origin of peripheral visual nervous 
impulses) has been laid. For some time past the minds of 
physiologists have been drawing near to the conception of a 
visual substance or of visual substances as the basis of vision. 
The speculations of Hering (Zur Lehre vom Lichtsinne. Wien. 
Sitzungsberichte, LXVI. — LXX.), though in their present form 
they may be spoken of as crude, and as perhaps raising more 
difficulties than they remove, have this great merit, that by 
the hypothesis of a visual substance (or visual substances) 
they bring visual (and by analogy other specific) sensations 
into the same category as ordinary nervous impulses and even 
muscular contractions (we might perhaps go so far as to say as 
protoplasmic molecular movements in general), and thus open 
up the way by which the phenomena of a visual sensation, a 
nervous impulse, the act of secretion, and a muscular spasm 
may be made to illustrate each other. 

Awaiting patiently the future discovery of the exact nature 
and relations of what may be called the true visual substances, 
we shall still be justified in maintaining that the visual purple 
is, if not directly at least indirectly, connected with vision; 
and, in this sense, I have not hesitated to retain the name 
' visual purple,' in spite of the fact that frogs appear to see 
perfectly well in spite of the absence of any store of it in 
their retinas, and that many animals which never at any time 
seem to possess such a store, have nevertheless very respectable 
vision; and I trust that an account in the English language 
of its behaviour and properties will prove not unacceptable to 
the physiologists and psychologists of this country. 

The translation has been made by Mrs Foster, who has 
found the task of converting Prof Kiihne's somewhat idio- 
matic German into readable English not free from difficulty; 
we trust, however, that his meaning has been in all cases 



Vlll PREFACE. 

correctly rendered, even if the diction should at times seem 
capable of improvement. I have gone carefully over the whole 
of the translation myself, and have added some notes in the 
form of an Appendix ; and Prof. Kiihne has himself kindly 
revised the proofs. 

M. FOSTER. 

Trinity College, Cambridge, 
July, 1878. 



PAET I. 

PHOTOCHEMISTRY OF THE RETINA. 



p. p. 



PAET I. 

PHOTOCHEMISTRY OF THE RETINA. 



In a recent communication to the Berlin Academy, Herr Fr. 
Boll announced the beautiful and, beyond doubt, important 
discovery, that the baciUary layer of the retina of all animals 
is in the living condition not colourless, as has been hitherto 
supposed, but of a purple red colour. During life, says Boll, 
the proper colour of the retina is continually being destroyed 
by the light which falls into the eye; it is restored in the 
dark, and after death only remains a few moments. Ani- 
mals which have been kept in the light are therefore very 
unsuitable for demonstrating the living colour of the retina, 
and the eyes of animals which have been exposed to the sun 
for a long time before death remain absolutely colourless. 
These facts illustrate the relation of the retinal colour to light 
on the one hand, and to vital conditions on the other. 

Whoever has busied himself with the retina will be reminded 
by Boll's discovery (and thereby receive a wholesome admoni- 
tion of the limits of his own ability), that he has already seen 
something of the kind before. He will perhaps remember 
that puzzling blood-clot — which at one moment he saw, or 
thought he saw, under the retina, and which the next moment 
disappeared. What he then passed over so lightly was nothing 
less than the key of the secret, how a nerve can be excited by 
light. In other words, it was the first fact disclosing the 
existence of photo-chemical processes in the retina. 

1—2 



4 PHOTOCHEMISTRY OF THE RETINA. [PART 

During my first attempts to test the truth of Boll's state- 
ments, I was under the impression, strengthened by what Boll 
had said, that the greatest haste was necessary in removing the 
eye, and taking out the retina; but I very soon convinced 
myself that a considerable time may be allowed for the ope- 
ration, since the visual purple, as I call the proper colour of 
the retina, may continue to exist for some time after the retina 
has ceased to be in a physiologically fresh condition, and even 
after death is only destroyed by light. Under the light 
of a good gas-lamp the retina may be quite leisurely spread 
out, and is then seen to lose its colour very gradually, for 
what takes place under bright daylight in half a minute, lasts 
in this case from 20 to 30 minutes, that is to say, much longer 
than Boll gives as the duration of tissue-life. And in the 
dark, or in the light of a sodium flame, the purple does not 
pass away at all, either in the frog or rabbit, at least not for 
24 or even 48 hours, in spite of decomposition evidently set- 
ting in. 

By the discovery of these facts, the path of experimentation 
with the visual purple is freed from many difEculties. The 
necessary preparations can be conducted leisurely in a dark 
chamber lighted by a sodium flame, and the object then brought 
out for examination into diffuse daylight. A room such as is 
used by photographers, into which the light passes through 
yellow glass or paper, answers the same purpose, but less 
perfectly. 

As we do not exactly know how long the rods or their 
parts remain alive after death, I treated the retinas of frogs in 
the sodium chamber with the most various reagents, many of 
which were without doubt capable of producing great structural 
alterations, to see whether the colouring and sensibility to light 
were thereby affected. The colour was destroyed by heating to 
100" C, by alcohol, by glacial acetic acid, by soda solution, both 
concentrated, and 10 per cent. It was not affected by a solu- 
tion of sodium chloride 0'5 per cent., by strong ammonia, by 
sodium carbonate solution, by saturated sodium chloride solu- 
tion, by alum, lead acetate, acetic acid 2 per cent., or tannic 
acid 2 per cent., by lying in glycerine,for 24 hours, by ether, or 



I.] PHOTOCHEMISTEY OF THE HETINA. 5 

by being dried upon a glass plate. In all these latter instances, 
the retina, when brought into daylight, was found to be still 
red, and bleached more or less rapidly, the purple passing after 
1 or 10 minutes into chamois colour', which gradually disap- 
peared, until there was scarcely anything of it visible. The 
depth of the colour naturally depended upon the other con- 
ditions of the retina, upon its being transparent, or cloudy. 
When it was opaque it gave an opportunity of proving the 
correctness of Boll's statement, that it is the outer layer, there- 
fore really the layer belonging to the rods, which is coloured; 
for an opaque retina is seen to be white in front, the red 
appearing at the back only. The colour comes out most beau- 
tifully after the action of ammonia, which makes the retina 
very transparent ; and the red of a retina treated with ammonia 
withstands the action of light from 10 to 20 times longer 
than an unchanged retina exposed to light of equal intensity. 
The dried membrane also retains its colour for a long time, 
but it also yields gradually to the light. 

From what has already been said concerning the methods 
of obtaining coloured retinas, it is obvious that all kinds of 
light are -not equally active in bleaching the visual purple. 
This is untouched by the photo-chemically inactive (or only 
slightly active) rays of the line D"^. It is only the very strongly 
coloured retinas (such as those of Rana temporaria) which shew 
any colour in sodium light, and then only a trace of it ; and 
sodium light is not quite monochromatic. The retina of a 
living rabbit examined with an ophthalmoscope in the approxi- 
mately monochromatic sodium light, appears bluish white, 
with something of a mother-of-pearl sheen, the vessels coming 
out with astonishing distinctness, like black lines drawn with 
ink. So also the pupil of an albino rabbit, when the eye is 
illuminated obliquely, looks quite black. Indeed the sodium 
light, which is so easily rendered sufficiently intense, may be 
expressly recommended as suitable for delicate ophthalmoscopic 
investigation. 

In order to ascertain which are the particular rays which 
bleach the visual purple, I placed retinas spread out on glass 
1 Se« Appendix, note. A. = See however p. 57. 



6 PHOTOCHEMISTRY OF THE RETINA. [PART 

plates, ia blackened moist chambers, covered them with glass 
slips, on which I had fastened strips of tin foil about one mm. 
broad, and then placed over them coloured plates of glass, or 
glass beakers filled with coloured solutions. For red I used 
blood solutions of such a strength that neither yellow nor 
orange were any longer visible in their absorption spectra ; also 
red glass plates, which, however, allowed some amount of purple 
as well to pass through. For blue I used ammoniacal cupric 
oxide, and for green, coloured glass plates, the spectrum of 
which consisted of only a narrow green band. Under the blood 
there was really no bleaching at all ; under the red glass the 
first trace of it appeared after 6 hours ; in the blue light the 
bleaching began after 2 hours, and in the green after from 
4 to 5 hours. Of course, owing to the intensity of the light 
under which these experiments were conducted not being very 
great, and also not really comparable, one cannot from them 
arrive at accurate conclusions concerning the problem ; but the 
apparently more powerful action of the more refractive rays, 
particularly of the blue light, was brought out clearly'. On 
removing the cover-glass from the bleached preparations, there 
was seen, where the retina was protected by the strips of foil, 
a beautiful band of unchanged purple, in other words a posi- 
tive photogram. I found the purple was as little changed by 
lithium light as by the blood-red, while, as might be expected, 
magnesium light very rapidly destroyed it. Once bleached, by 
whatever means, the purple never returned, either in the dark, 
or in different coloured light, or by being exposed to heat, or 
by the ultra red rays coming through a smoked glass illumina- 
ted by the sun^ 

After I had followed out the experiments I have described, 
with animals which had, according to Boll's directions, been 
kept in the dark, I was anxious to see how the retina of a frog 
would appear, which, immediately after exposure of the eye to 
light while the animal was still alive, had been prepared as 
rapidly as possible in a chamber of yellow light. According to 
Boll's opinion, I expected to see it bleached to a perceptible 
degree ; I found it on the contrary as red as others. It is 
' See Appendix, note B. ^ See Appendix, note C. 



I.] PHOTOCHEMISTRY OF THE RETINA. 7 

therefore unnecessary to keep the animal in the dark just pre- 
vious to the experiment. As the daylight was not very bright, 
the sky being cloudy, though quite sufficient for microscope 
work, I tried illumination with magnesium light, but that also 
was followed by the same result. I therefore came to the con- 
clusion that Boll was mistaken in ascribing the failure that 
he once had in demonstrating the presence of the purple, to 
the fact that the frog had been kept in the light ; if he found 
his preparations equally bleached, this must have been due to 
exposure to light during the time they were being manipulated. 
To discover why it was that the visual purple remained 
thus apparently unchanged in the physiological act of vision, I 
isolated the retina of a frog and laid it on a glass plate, leaving 
the other in the bulb, which I had however extirpated and 
had opened wide by an incision along the equator. Both pre- 
parations were then brought into daylight which was still not 
very bright, and were left there until the first was perfectly 
bleached ; the second one was then brought back into a chamber 
lighted by sodium light, the retina taken out, laid upon glass, 
and brought out again into ordinary light ; it was at first of 
a deep red, but quickly became colourless. The sky not being 
clear, I conducted the same experiments with a magnesium 
lamp, and always found that the visual purple was restored so 
long as the retina remained in the eye lying upon the 
choroid, though it was protected from light and air by nothing 
more than thin capillary layers of the vitreous humour. On 
the following day, when the mid-day sun was almost clear, 
and shining so brightly that I was unable to look up at it, I 
repeated the experiment, by bisecting a frog's eye, removing 
the vitreous humour, and exposing the posterior half containing 
the retina to the sun for four minutes. I found even then red 
specks in the chamois-coloured retina, only the edges being 
completely bleached. In an entire bulb which I had exposed 
properly arranged to the same sunshine for 25 minutes, I still 
found very faint red spots in the midst of a great deal of 
chamois colour, but this was probably owing to the pupil be- 
coming very contracted during the illumination. Since I 
always prepared the retina in sodium light in these experi- 



8 PHOTOCHEMISTKY OF THE EETINA. [PAET 

ments, it might be thought that the period of photo-chemical 
rest thus occasioned, although short, in some way or other 
caused the return of the purple. This however is not the case, 
for if a divided eye be held up against the daylight for a period 
quite sufficient to bleach an isolated retina, and then, while the 
light is still falling on it, the retina be rapidly drawn out of 
the bulb, this will always be found to be at the first moment 
beautifully red. It is seen that my observations support those 
of Boll in this limited sense, that the retina may be bleached 
in living animals, after long continued exposure to direct 
sunlight ; I may moreover add that frogs which had been left 
to themselves for several days, in a glass case, in a very sunny 
place, at the end of that time had colourless retinas. While 
I do not agree with Boll's experience of the effects of "a 
moderately lighted chamber," we are at one as to the results 
of more intense illumination. 

If the photo-chemical processes which take place in the 
retina separated from the eye, are taken as representing those 
which are going on in the living eye, the visual purple may be 
conceived of as continually being destroyed by the act of seeing, 
and by some process or other being as continually renewed; 
and indeed Boll has expressed some such view as this. The 
oculist led by experience would immediately seek for the pro- 
cess of regeneration in the nourishment brought by the circu- 
lating blood ; for this is a favourite way of accounting for most 
of these kinds of events. In the case in point, however, the 
matter is far less complicated. 

That which restores the visual purple is something nearer 
to hand, and cannot, in a frog's eye at all events, depend on the 
constant renewal of the blood, since an eye, when taken out 
and opened, exhibits the same apparent indifference to light as 
when connected with the whole body and the nutritive currents. 
If therefore we are correct in supposing the sensitive purple to 
be continually restored, the regeneration must proceed from 
something lying behind or on the rods, that is, from the retinal 
epithelium or from the choroid. Something must be there 
which either prevents the purple becoming lost, or re-creates 



I.] PHOTOCHEMISTRY OF THE RETINA. 9 

it. The idea at once suggested itself that the mere pigment 
as pigment had something to do with the matter, because a 
more intense action of light is to be expected if the retina, 
which usually only receives light from the front, is in addition 
lighted from behind, as is the case when it is spread out upon 
a white surface, instead of lying in its natural condition upon a 
velvety black ground. It could not however be supposed that 
this would protect it so long and so completely as is found to 
be the case. Moreover, I could not discover that spreading out 
the retina with the rods turned downwards upon a dead black 
surface had much effect upon the time of bleaching ; and the 
following experiments will perhaps shew that the evident con- 
tinual renewal of the sensitive colour is due to something else 
than the pigment, which, as is well known, is not present in 
albinos, and in many animals lies behind a tapetum. 

As a proof that it is the choroid, including the retinal 
epithelium, which alone protects the purple from bleaching in 
light, it is sufficient to remove the retina in such a way, that a 
number of black shreds of the choroid still adhere to it, and 
then to spread it out on a thin glass slip, and expose it on 
every side. The manipulation for this purpose is not difficult, 
if the bulb is so cut out tJiat a hole is left at the entrance of 
the optic nerve, and consequently the spot got rid of which 
offers resistance to the removal of the retinal membrane. From 
a bulb so manipulated, it is very easy to get a retina spread 
out without a crease, especially if a few meridian incisions are 
made in it. These trivial matters are important, because in 
badly prepared and creased retinas the pigment actually pre- 
vents the access of light to certain parts of the retina. If now 
the dark strips are taken off the fully bleached preparation, 
the parts which lie under them will be found to be very deeply 
coloured. 

Another experiment, which demonstrates the same thing, 
consists in tearing the bisected bulb slightly until as usual 
some folds of the retina bulge out, then letting the light shine 
in, and quickly pulling out the whole retina. Where the folds 
were, white stripes will be found, all the rest remaining red. 
I also made the following experiment. Having made an 



10 PHOTOCHEMISTRY OF THE RETINA. [P^KT 

equatorial incision, I seized the retina along a considerable 
portion of its edge, and very carefully lifted it up from the 
pigment layer over a good half of its area. I then slipped a 
thin morsel of porcelain under the raised portion, and exposed 
the whole to daylight, until it was completely bleached. Of 
course the bleaching can only be ascertained with certainty 
in the flap which is raised, as the visual purple cannot he 
recognized in the dark reflecting hollow fundus of the eye. I 
next allowed the bleached retina to sink slowly back in sodium 
light on to its natural support, and to lie for some minutes 
upon it, making sure the meanwhile that my manipulation had 
been successful in avoiding the formation of creases; and then 
I drew the entire retina away. It was uniformly red all over, 
and one could not even discover a line to mark the limit of the 
half which had been lifted up. A retina therefore bleached 
by light may regain its purple colour by simple contact with 
its natural back-ground. Still it remained to make the whole 

O 

experiment in active light. This also succeeded, but the half 
which was restored was somewhat paler than the other. I have 
no doubt of the success of this operation undertaken by anyone, 
and I even go a step farther and recommend the cutting out of 
a flap, the bleaching of it upon a plate and the laying it back 
again upon the exposed pigment, by which it will be seen that 
any piece replaced in a normal manner always recovers its 
purple again. The regeneration has so often and so well 
succeeded with me by this method, that I was seriously led 
to try with a piece of tissue paper whether the cup of the eye 
did not contain a small quantity of red secretion; the morsel 
of tissue paper however came out moist it is true, but quite 
colourless. 

In a frog's eye such experiments carried out with necessary 
care need not be hurried ; nevertheless the regeneration of the 
purple, unlike the mere continuance of the colour or the main- 
tenance of its sensitiveness to light, is dependent on the action 
of living tissues, and when the tissues cease to be alive the 
regeneration can no longer take place. I have placed the eyes 
of frogs in Oo per cent. NaCI at 43" C. for 10 minutes, then 
bisected them, exposed them to the light, and always then 



I.] PHOTOCHEMISTRY OF THE RETINA. 11 

found the retinas colourless. Since the retinas of the eyes so 
heated were still red before exposure to light, they must have 
been bleached by the light. The same thing occurred in eyes 
which had died after being kept a day at about 20" C. It may 
be here mentioned that these failures in dead eyes prove 
that the pigment, considered as mere pigment and not as a 
living tissue, is ineffectual as regards the maintenance of the 
visual purple. 

If the regeneration of the purple depends on a continuance 
of vital activity in the pigmentary support of the rods, it is 
obvious that the rapidly decomposing organs of mammals are 
ill suited for these experiments. Here rapidity of manipulation 
is above all things necessary, and yet I have often succeeded 
in drawing off from the hinder half of the eye of a rabbit, 
beautifully red pieces of the retina, even after it had been 
exposed to the light for two minutes, that is for a time quite 
sufificient to bleach an isolated piece until the only colour left 
is that belonging to the blood-vessels. Still even in an albino 
rabbit, where the circumstances must be allowed to be pecu- 
liarly favourable, I think I have been able to distinguish 
between the colouring of a piece of retina which has remained 
in its natural position, from that of one which had been re- 
moved, especially when the former, after the bleaching of the 
latter, had been spread in the same way on porcelain. Never- 
theless I cannot speak with certainty on this point, since the 
retinas of the specimens of albinos at my command for this 
demonstration were taken from a variety which just now is 
very difficult to get here, and in spite of being kept long in 
the dark failed to shew a really good intense purple, and after 
being exposed to light exhibited only a slight variable pale 
orange colour, which indeed is not an altogether unknown fea- 
ture in mammalian retinas. 

It would be especially interesting to investigate this orange 
colour*, which perhaps co-exists in company with the purple, 
since Boll has made the very important observation, that blue 
rods are very frequently present in the retina of the frog. 
That there are many albino eyes possessing a well-developed 

1 See Appendix, note B. 



12 PHOTOCHEMISTRY OF THE RETINA. [PART I. 

purple I saw in later experiments, of which I will give an 
account on some future occasion. 

The last-named series of experiments lead me back to the 
conclusion that the maintenance after death of the vital con- 
ditions in the outer layer of the visual apparatus is to be 
recognized, not by the existence of the visual purple, and its 
destruction by light, but by its relative indifference towards 
light, and I think that the being able to detect this reaction 
in the frog's eye for so long a time after death shews a happy 
agreement with Holmgren's and McKendrick and Dewar's 
beautiful experiments upon the galvanic currents of the retina, 
and their changes during stimulation by light. 

What part of the choroid serves to regenerate the purple 
can for the present only be conjectured'; probably it is not 
to be sought so much in the dermic layer of the choroid as in 
the epithelium in which the rods are embedded, and which has 
been rightly considered as part of the retina. The retina, so 
long as it is maintained in its natural connections with this 
epithelium, resembles not so much -a photographic plate as a 
whole photographic workshop, in which the operator, by bring- 
ing new sensitive material, is always renewing the plates, and 
at the same time washing out the old image. 

^ See Part II, and Appendix, notes D and F. 



PAET 11. 

VISUAL PURPLE. 



PART II. 

VISUAL PURPLE. 



The conjecture which was expressed in the introduction to 
the former treatise, that the red colour of the retina had been 
seen by many observers long before it had obtained so great 
an interest through Boll's communications, may be readily 
confirmed by studying the rich literature on the structure of 
the retina. Apart from A. Krohn's discovery of the red rods 
in the Cephalopoda quoted by Boll, their occurrence in verte- 
brates was first suspected by H. Miiller in 1851. The im- 
portance of the subject will justify a verbatim quotation of the 
essential statements. 

A. Krohn says in his paper delivered on Sept. 4th, 1839', 
"Close by the dark stripes the fibres (visual rods) are of a 
reddish colour, but at their ends turned towards the vitreous 
membrane quite colourless ; the transparency and the rose-red 
sheen of the inner surface of the retina are thus accoxxnted for." 

This is confirmed by Hensen in p. 39 of his paper on 
the eyes of Cephalopoda" in the following words : " In the 
fresh retina they (the rods) possess red, shining, homogeneous 
contents." 

M. Schultze remarks in speaking of the eyes of Cephalopoda 
in 1869^ : "The rose-red rolour depends upon a diffuse coloura- 
tion of the entire thickness of the bacillary layer (Plate l„ Fig. 

' Printed in the Verhandlungen der Leap. Carol, Akad. Bd. xix. 11, 1842, 
" Supplemental Observations on the Structure of the Eyes of Cephalopoda." 
' Leipzig, 1865. 
' Archiv f. Mikrosc. Anatomie, Bd. ii. S. 3. 



16 VISUAL PURPLE. [PART 

1, coloured drawing), but is however only visible in fresh 
specimens, as already mentioned by Krohn. Under the micro- 
scope it is recognizable only in thick masses of separated rods." 

Schultze further remarks that he has seen the most 
beautiful rose-red in a large specimen of Loligo almost entirely 
free from pigment. Well-known, moreover, are the descriptions 
and beautiful figures which the same great authority on the 
retina has given in his memoir on the compound eyes of 
Crustacea and insects \ in which he treats of the gigantic purple 
coloured rods of these animals. 

As far as vertebrates are concerned the first accounts are 
found in the epoch-making memoirs of Heinrich Muller. In 
1851" Muller writes: "The rods of frogs when lying thickly 
together appear to be in themselves of a somewhat red colour, 
and an isolated rod may be seen to be alternately colourless 
and coloured according as it is seen sideways or on end. In 
1856" Miiller returns to this: "The substance of rods is often 
seen to be reddish, as I have remarked in my first notice, if it 
is of sufficient thickness ; as for instance when a rod is viewed 
on end, or several are seen lying together, one over another. 
This colouring is not uniform all over, but is sometimes 
stronger, sometimes weaker, often unobservable ; and although 
it appears in eyes which are quite fresh, it may perhaps depend 
on an imbibition of the colouring matter of blood. Also the 
colouring which appears in the cones of birds is diffused by im- 
bibition over the surrounding parts." 

Six years later Fr. Ley dig* writes : "The rods of Am- 
phibia (Rana, Pelobates for example), if they lie in great 
numbers together, possess a rose-red colour; in the case of 
many fishes, for example Cobitis fossilis, they have a yellowish 
sheen. The fresh retina of the frog appears to the naked eye 
to be of a lively, satiny red." 

After this it is not till 1866 that the red rods of the verte- 
brata are spoken of This time again by M. Schultze (in Vol, 

1 Bonn, 1868. 

" Zeitschriftf. Wiss. Zoologie, iii. S. 234 — 237. 

3 Ibid. vm. S. 1—122. 

^ Lehrb. d. Histologic, 1857, S. 238, 239. 



II.] VISUAL PUEPLE. 17 

II. of his Archiv, p. 199), who says : " Remarkably long rods are 
seen in the rat, the freshly removed retina of which, spread out 
with the choroidal surface uppermost, shews a strikingly marked 
satiny appearance, with a red sheen similar to the retina of the 
owl and the frog." In the same work, p. 208, he further says : 
" This (the retina of the owl) exhibits in a very remarkable 
degree the red satiny sheen, which is present in those mam- 
mals in which the rods are of an unusual length." 

It will be seen from these extracts that the investigators 
lay most weight upon the fresh condition of the retina, without 
thinking of its relations to light, and that the colouring is 
in their view connected in part with the length of the rods ; 
while the repeated mention of the satiny appearance and the 
glittering colours gives rise to the supposition in the minds 
of most investigators that the appearance depends rather upon 
interference than upon actual pigment. 

A similar view is directly enunciated by Boll, who, while 
considering colouring as a direct indication of tissue life, pro- 
poses to investigate whether it is caused by interference or by 
a pigment. 

Without wishing to say that the literature on the retina 
offers us no further account of the red colour of the rods, I will 
conclude this brief notice by referring to the memoir of E. Rose 
upon the action of santonin, in which he speaks of a red and 
greyish-red colouring of the retina of the rabbit, which I take 
as referring to the visual purple \ 

In continuing my investigation on the visual purple, I do 
not specially aim in this work at giving a systematic review of 
its distribution in the animal kingdom ; still less so, since the 
facts which have been given leave no doubt of its frequent 
occurrence ; a large number of invertebrates possess red rods^, and 
among vertebrates the colour has been recognized in osseous 
and cartilaginous fishes, in amphibia, in birds', and in mammals. 

With the discovery of the sensitiveness to light of the visual 

^ Virchow's Archiv, Bd. xviii. 15, 16. 

" It would appear, however, that the pigment of the invertehrate retina, 
being stable and not sensitive to light, must be something quite different from 
visual purple. Eiihne, Untersuch. Bd. i. Hft. 4, p. 456. 

' Boll says it is present in the pigeon, but see p. 26. 

F. P. 2 



18 VISUAL PUEPLE. [PART 

purple, the idea has very commonly arisen, as I gather from 
numerous letters and other utterances, that we are now in a posi- 
tion to understand tolerably well how the excitation of the optic 
nerve is brought about by the action of light. I can to a very 
limited extent only agree to this opinion, although for the 
purpose of investigation I feel obliged to put before myself the 
hypothesis, which indeed forces itself on every one — that the 
various decomposition products of the visual purple, namely, 
the orange, the yellow, and in particular the colourless sub- 
stances, serve as chemical stimuli for the ends of the optic 
nerve, while the original visual purple serves as an inert 
medium having no effect upon them. I would not however 
venture to make this hypothesis without joining to it the 
caution that the visual purple must not be looked upon as the 
only substance in the retina which is sensitive to light. The 
idea that the movement of the luminous aether is transformed in 
the retina into chemical processes, is one which has been float- 
ing about for many years, even when there was no suspicion of 
the existence of the visual purple, and there was nothing to 
justify the assumption that the substances necessary for such 
an action might be recognised by their colour. It must be looked 
upon as peculiarly fortunate that one of these is recognizable by 
this quality. It must first be proved that there is blindness 
where the visual purple is blanched, and it still remains to be 
proved that all visual organs are provided with the purple. A 
second supposition which has met with much acceptance concerns 
the significance of the products of bleaching, in connection 
with the question as to what are the structural elements of the 
retina which are first laid hold of by the light ; in other words, 
what are the real terminations of the optic fibres. It is easy to 
imagine that the finest radiating processes of the simple con- 
ducting fibres embrace the refractive bodies in the inner limbs 
of the rods, and spread out in the red outer linib. In that case 
the final terminations, which might continue to be considered 
similar to the fibres of the optic trunk and indeed to all genuine 
nerves in respect to structure, composition, and irritability, 
would be in the most simple manner stimulated, throuo^h the 
medium in which they lie becoming loaded by photo-chemical 



II.] VISUAL PURPLE. 19 

action with some caustic agent. If this view is correct, one 
would imagine that the posterior surface of the retina would 
in the moment of illumination serve as a stimulus to the fresh 
transverse section of the motor nerve of a frog. Since the first 
day that I began to be occupied with the visual purple I have 
often made such an experiment with the most sensitive pre- 
parations, and under circumstances varying as much as pos- 
sible in respect to the duration and intensity of the illumina- 
tion ; but I have never seen any contraction of the frog's limb take 
place, not even when a dazzling ray of sunlight was suddenly 
allowed to fall into the dark chamber upon the retina which was 
laid over the nerve. As it may be doubted if a transverse section 
of a nerve, in which the medullary sheath might easily cover 
up the axis cylinder, is as sensitive to the action of a chemical 
stimulus as the very delicate fibres which possibly exist in the 
rods may be, I have tried the bleaching of the retina upon 
the outer as well as the inner surface of the skin of a frog 
prepared for exhibiting reflex action, but always without any 
result ; that is to say, without any irritation of the cutaneous 
nerves being manifested by reflex acts. I do not consider this 
hypothesis disproved by such experiments, against which many 
and obvious objections, which need not be at present dis- 
cussed, may be raised. I am, however, of opinion that one 
ought to give the preference to another view. 

When we examine nerves which are sensitive to chemical 
stimulation, it is always found in every case that the con- 
ducting fibres terminate in peculiar epithelial organs ; and that 
these latter differ from ordinary nerve-fibres in respect to 
structure, composition and irritability can be doubted by no 
one. Such is the structure, pwr excellence, of the nerves of 
smell and taste ; and such a sensory epithelium is not to be 
regarded as a mere envelope for the nerve-fibrillse which pass 
through it and project over the surface like hairs or bristles. If 
this were the case we might predicate similar sensitiveness where 
these terminal fibrillffi were found, and we should expect accord- 
ingly to feel a burning in the eye by means of Cohnheim's 
nerves in the epithelium of the cornea, at the same time that 
we perceive musk or oil of roses by the nose. Shall we then 

2—2 



20 VISUAL PURPLE. [paut 

make an exception in the case of the terminations of the optic 
nerve and its sensory epithelium, in the day when we find in 
it substances which are chemically changed by the action of 
light, and recognize in the excitation by light a chemical stimu- 
lation? Certainly not. The rods appear to us now physio- 
logically also as sensory epithelium, like olfactory and gustatory 
cells; their cuticula, that is the outer limbs, being the part 
which is decomposable by light, while the nucleated proto- 
plasmic inner limbs appear as that part which bleaching pro- 
ducts throw into excitation. By this theory there is room left for 
the possibility of a thread of the body of the cell, or a process 
of semi-fluid matter containing such a thread, stretching far into 
the termination of the outer limb, as a soft Ritter's fibre, or 
as the core of a Hensen's canal. As the olfactory cells are 
changed and stimulated by the most minute quantities of odorcus 
substance, so minute as to defy calculation, so we may suppose 
can the visual cells be acted upon by the least trace of the 
bleaching substances which reach them. This conception of the 
origin of visual sensations affords at the same time a desirable 
link in the doctrine of the development of the higher sensory 
organs, which, in the case of the retina, has been clearly and 
logically carried out in Schwalbe's' exposition of the subject. 
Whoever is acquainted with the minute structure of the retina 
will here ask the question, what signification under such a view 
can be attached to those structures which were investigated with 
such care by M. Schultze and described by him as fibres in 
the furrows of the surface of the rods, and as the thread appara- 
tus and the fibre-baskets around the inner limbs ? In reference 
to this it must not be forgotten that the retinal epithelium of 
the choroid has been proved to be a very important physiological 
or chemical constituent of the retina, or, stated shortly, as a 
purpurogenous gland, the cells of which can scarcely fail to pos- 
sess a very peculiar complicated innervation. Now I do not see 
that irritable fibres can arise from other source than from the 
nervous mass of the retina. And since one knows that the pro- 
jections of the epithelium cells, which Czerny'ten years ago 

1 Handbuch der Ophthalm. v. Orafe u. Saemisch. 
' Wien. A kad. Ber. LVi. 



n.] VISUAL PUEPLE. 21 

suspected to be pseudopodic processes, are structures in the high- 
est degree variable (in which, for example, the pigment during 
life changes its place in a most striking manner, and arranges 
itself in layers), it may easily be believed that not only the thread 
apparatus, the thread baskets, and upper fibres, but even those 
needles of M. Schultze's passing through the external limit- 
ing membrane are in part something more than mere cuticular 
thickenings, or cement material — are indeed very fine nerve- 
fibrillse. In the last place, I must not allow still another hypo- 
thesis to be passed over, according to which the changes in the 
visual purple may be the result of, and not, as we have been 
supposing, the cause of an irritation by light of specific nervous 
elements. The sensitiveness to light of all dead retinas clearly 
enough contradicts this ; but before I was aware of that, I had 
not omitted treating fresh retinas of frogs in the dark with every 
kind of electric stimulus ; the consequence was, as every one 
could now predict, entirely negative : the visual purple was never 
bleached. Not to pass over anything, it may here be remarked 
that one might conceive of the terminal organ of the optic nerve 
exhibiting sensitiveness to light by means of substances sensitive 
perhaps to light and yet • colourless, which on their part first 
destroy the purple, — we might suppose this if the unchangeable- 
ness of this latter towards the most varied chemical influences 
was not already known. This does not exclude the importance 
of experiments which would prove alongside of the bleaching 
other chemical decompositions in the retina by light; only I 
myself have not been fortunate with these, in so far at least as 
the simplest means, namely, the litmus reaction, has never 
shewn any change of the alkalescence. A fresh retina of a frog 
is, after the most thorough washing away of the vitreous 
humour in '5 per cent, sodium chloride, clearly alkaline, and when 
pressed on litmus paper orLiebreich's tablets leaves a distinct 
blue mark. If the contrary appearance comes out, this is 
caused by the red rods being drawn into the pores and rough- 
nesses of the paper ; when the reaction is confined to a circle 
surrounded by the fluid, the clearest blue is never missing, and 
it never will be seen passing into red in the light, while when 
the red masses of rods obscure the reaction the blue will be 



22 VISUAL PURPLE. [PART 

recognized as coming afterwards in the light. If I mashed up 
a retina in an agate mortar and then let the mass bleach in the 
light, this still gave an alkaline reaction. 



The foregoing considerations caused me in the next place 
to examine whether all those elements of the retina which up 
to this time had been considered as sensitive to light contained 
the visual purple. It immediately struck me that I had 
never recognized a trace of purple in the cones of the frog 
In the retina of a frog which has been smoothly spread out, 
with the choroidal surface in contact with the cover-glass of a 
flat moist chamber, the cones, as is well known, may by 
proper focussing be readily seen between the rods, in the 
deeper layers where they are specially recognizable by the 
highly refractive bodies of their inner limbs. I have never 
been able to discover here any red colouration, and with a 
higher focus the space between the rods is never found other- 
wise than complementary to the visual purple, that is to say 
blue-green, just as it is in those spaces between the rods in 
which the cones are absent. This blue-green colour is visible 
also, as Boll has already mentioned, in a certain number of the 
rods, namely, in such as are somewhat turbid ; and which 
generally, unlike their clear red neighbours, do not, when an 
object is slipped in between the mirror and the diaphragm 
of a microscope, shew the Leeuwenhoek images, which Boll 
and M. Schultze found reflected from the rods of the frog. 
This green colouration toning down into grey which, arising first 
from simultaneous, is increased by successive contrast, will 
invariably be seen even in the rents and holes of the prepara- 
tion, provided these are not too large, and particularly if they 
are filled with some cloudy substance which obstructs the 
perfect passage of the light. And in cases where the rods 
are lying down and scattered about like the ears of com in 
a cornfield which has been beaten down by wind and rain, 
(a state of things which by the naked eye is recognized in the 
fresh red retina by the occurrence of shiny satiny streaks) this 
is seen in the microscopic view as large streaks and tracts of a 



II-] VISUAL PURPLE. 23 

blue-green colour. The origin of these striking appearances is 
that the rods are not sufficiently red to allow of the purple 
being recognized except in the direction of their long axis. 
One must be very quick not to lose the time of the most 
intense colouration, if one wishes to recognize the colour in 
the isolated rods lying on their sides. Perhaps moreover besides 
the want of depth in the transverse diameter, the reflection 
of light from the outer surface of the rod is unfavourable to 
the perception of the colour. It will be understood from this 
why the turned-up edge of a preparation of the retina only 
appears red when the rods lie one over another in sufficient 
numbers ; and why the circles and lines of the turned-up rods, 
which without the red surroundings appear to be grey, must 
appear a bluish-green by contrast when seen between those 
which are recognizable as being red. That the rods which 
remain standing, if they appear to possess the blue-green colour, 
are in reality not coloured, may be proved by inserting a disk 
of black cardboard with fine holes pierced in it in the place 
of the ocular micrometer. The red ground is then covered, 
and by turning round the object on the slide, one can bring 
into view at one time red and at another colourless grey, 
less refractive transverse optical sections of the rods. If all 
the rods are bleached, still those which are the complemen- 
tary ones may be always clearly recognized, although their 
colour has of course disappeared, by their inferior transparency. 
It is quite different with the really green, that is, strictly grass 
green rods — which are present in the retina of the , frog. 
These are mostly transparent, they give the Leeuwenhoek 
images, they bleach somewhat slower than the red, they 
stand the isolation test of the ocular diaphragm; and when 
I have occasionally met with them isolated, and standing 
on end, their colour has been fully developed. I do not doubt 
that Boll had these rods particularly in mind in his second 
communication to the Acad. d. Lincei of 7 Jan. 1877 S. 3, 
4. Notwithstanding this, I am unable to bring their occur- 
rence into relationship with the colour of the light to which 
the eyes had been exposed : a relationship which Boll sus- 
pected after making experiments upon frogs which had lived 



24 VISUAL PURPLE. [PART 

under green glass. In frogs which had been kept in the dark 
I found this kind of rod also inconstant. 

As the outer limbs of the cones in the retina of the frog 
are very short and come to a fine point, it is possible that 
a purple colouration, even if present, might not be recog- 
nized. I have sought evidence of this in the retinas of birds 
(pigeons and fowls), which possess a large number of cones ; 
although here I had from the outset to expect difficulties of 
observation on account of the well-known pigment-globules, 
and other colourations of the inner limb of the cones. The 
result of my numerous observations was negative : generally 
there was no visual purple at all to be recognized in the retinas 
of these birds. It was impossible for my sense of colour, as 
well as that of all other persons before whom I laid the fresh 
retina, to perceive any change of colour in diffuse day or direct 
sunlight — either in the central portion which in daylight ap- 
pears to be chiefly coloured by red pigment-globules or in the 
peripheral greenish-yellow coloured portions. Since hoth regions 
contain rods as well as cones, and since at the periphery most 
elements are free from pigment-globules, I suppose one ought to 
see the visual purple of birds if it exists. Where in fresh prepara- 
tions heaps of separated outer limbs lay in thick layers together, 
I have never seen any indication of purple. Therefore pigeons 
and fowls have no coloured substance in the outer limb which 
can be changed by light ; and this proves that sight does 
not entirely depend upon the presence of such a substance. 
From M. Schultze's discoveries upon the eyes of nocturnal 
birds, we know that intensely coloured pigment-globules are 
not present in the retinas of all birds ; and if doubts are found cast 
here and there upon this important statement of the morphology 
of the retina, this is most probably due to the modest statement 
of the great histologist, who does not pronounce decidedly on this 
point. The doubts of many observers may be due to the fact ' 
that they are unable to reconcile the observation of the absence 
of red globules in the owl's eyes, the globules in this animal 
being at the most pale yellow, with the mention of a strongly 
marked red colour in the rods, an appearance which previous to 
the discovery of visual purple had not been understood. At the 



!!•] VISUAL PURPLE. 25 

present moment this latter circumstance, which is quite casually- 
related by Schultze, is of peculiar importance; and I have 
not doubted for a moment that it is correct, and that it in- 
dicates a further fundamental difference in the nature of the 
retinas of day and night birds. 

Having obtained a living owl [Striai passerina, s. Athene 
noctua, Retzius), I am in a position to confirm Schultze's' dis- 
covery. The bird was brought to me in a very lively state early 
in the morning, and kept for four hours in the dark, before being 
decapitated. On opening the eye the retina remained attached 
to the vitreous humour free from pigment — so that it was 
cut away from the entrance of the optic nerve without any in- 
jury. It shewed strong absorption in sodium light and appeared 
quite grey in it ; this arose from the unwontedly intense 
purple (almost passing into blue) which spread uniformly over ■ 
the entire posterior surface. I found, as M. Schultze said, 
that the rods were longer than in all other vertebrates, and 
not very narrow. Between the rods there were in many places 
cones visible with delicate less shiny outer limbs, which could 
not be called short, though they certainly only reached to one 
quarter of the length of the rods. The refractive bodies at the 
junctions of the inner and outer limbs were colourless. The 
purple did not bleach in light more rapidly than in other animals, 
during which it passed into a somewhat persistent orange, and 
here, as in all cases, was present only in the rods. When 
M. Schultze stated that in the owl the yellow cone-globules 
are only less intensely coloured than in other animals and not 
entirely absent, it being only the red ones which are lack- 
ing, I cannot help thinking that in making this statement (and 
we must remember that he was treating the subject in a very 
general manner only) he was influenced by having seen the 
orange of the entire retina — a phenomenon which could hardly 
have escaped his notice. His accounts and figures of Sti'ix 
aluco and Strix noctua are too positive for me not to recog- 
nize that they distinctly differ from mine ; for, as I have said, 
I have never been able in any species to discover any yellow 
colour in the globules in question ; as soon as the orange was 
1 Archiv f. Mikroscop. Anat, ii. 



26 VISUAL PUEPLE. [PAET 

bleached; which took place in about 45 minutes in imperfect 
afternoon light, not only the whole retina appeared quite colour- 
less to the naked eye but also the globules transmitted no 
yellowish tinge even when their surroundings were no longer 
coloured. In the black pigment epithelium whose cells are 
provided at the side of the rod with a tuft of remarkably 
long processes, I found no transparent pigment-globules. 

In another owl, Aluco stridua {Syrnium aluco, Linn.), I 
found the retina, which tore close by the pecten and came 
away with the vitreous humour free from pigment, provided 
with fine rods which were somewhat less long and corre- 
spondingly appeared weaker, and less uniform with regard 
to the purple colour. The purple in which a violet tone 
was very prominent, passed when exposed to light into a 
pale chamois before it became entirely bleached, never be- 
coming orange or pure yellow. Between the rods lay a fair 
number of cones, of which the greater number were provided 
with colourless, or at the utmost very extremely faint yellow 
globules, while a smaller number were distinctly yellow or orange 
coloured, and a very few isolated ones were of a red colour. 

In the retina of a tower falcon (Tinnunculus alandarius, 
Brisson) I found, contrary to my expectations after the ex- 
perience of the absence of the purple in the pigeon and the 
fowl, richly purple coloured rods again present ; but these 
together with their quite intense violet colour were arranged 
in stripes and spots in a still more striking manner than in 
the Aluco, and confined to those places where there were few 
cones or only such as had colourless or very feebly coloured 
globules. Where cones with red yellow or greenish yellow 
globules were found in any number, the surrounding rods were 
colourless. Unfortunately I could not sufficiently early ex- 
amine the fovese of the retina of the falcon, so that I can give 
no opinion upon the presence or absence of the purple in 
these important spots. As Boll decides upon the presence 
of the purple in the eyes of pigeons, I have striven by increasing 
the material to arrive at greater certainty. My hopes that the 
albino pigeon, with the deep ruby red pupil, perhaps might 
be without the disturbing pigment -globules were not fulfilled ; 



II.] VISUAL PUEPLE. 27 

for I found this retina did not differ from that of other pigeons. 
The cones contained yellow, greenish yellow and red pigment- 
globules, the rods had no perceptible trace of purple. As in 
the retina of all pigeons there was often the appearance of a pale 
flesh-coloured colouring in the mosaic of the rods which lay 
between remarkably ruby red cones ; yet this was seen only in 
those places where the preparation was not quite smooth and 
the persistency of the coloured sheen even in direct sunlight 
excluded the supposition that the tint was due to visual pur- 
ple. According to all appearance, the visual purple in the 
bird's eye is deficient in proportion as the retina is provided 
with other more stable means of absorbing coloured light, I 
mean the coloured globules of the cones. This is at least the 
case in the nocturnal and predaceous birds, and is entirely true 
as regards the pigeon and the fowl. 

Besides the owl, I have been able to make investigations 
concerning the purple upon another species of nocturnal ani- 
mal. In the bat (Bhinolophus hipposideros, Bechst), which 
is plentiful here, I tried to make sure of the colour of the 
retina. Unfortunately the eye of this species is so small that 
one must be content to cut it in pieces and to separate it 
out upon the microscope slide ; this I naturally carried out 
with sodium light — a treatment which the reader must pre- 
suppose in all the experiments which are described in this 
treatise, where the contrary is not expressed — using animals 
which had been kept in the dark. I have gone further, and killed 
those animals in which I found no visual purple, after they 
had been kept a long time in the dark with a bandage over 
their heads, and I took in hand the preparation of their eyes 
with the greatest haste, either in the minimum of light which 
sufficed for the operation or with a blue hght from an ammonia- 
cal solution of oxyde of copper whose freedom from red, yellow 
and green rays had been proved by the spectrum, and with 
which one worked scarcely more conveniently than in darkness. 
I took these precautions because one could not teU whether 
there might not be present visual pigments far more sensitive 
or more affected by light of lower refrangibility, than is the, 
visual purple. In spite of all these precautions, I have never 



28 VISUAL PURPLE. [PART 

been able to perceive even slight indications of red colouring 
in the rods of the bat, which are certainly very minute, and 
which, as is well known, do not surround any cones. I must 
therefore maintain that the rods in this species which I have 
examined are colourless, even if colouration should really be 
proved to exist in other species of this great order where the 
rods are longer. 

"Without wishing to say that genuine rods which are not 
red are altogether useless for vision, it appears to me doubtful 
on account of their mere existence in the bat to infer a power 
of sight, at all events one at all comparable to our own 
sense ; since we know from Spallanzani's famous experiment, 
which Briicke in his " Vorlesungen iiber Physiologie " justly 
interprets from the stand-point of the knowledge of the 
present day, that these animals shew no awkwardness in their 
movements after their eyes have been destroyed. With the 
same right that we conclude from the behaviour of blinded 
bats that they possess an astonishingly fine sense of touch 
and sensitiveness to rays of heat and the like, we may also, 
conclude, that they are certainly not accustomed to sight in 
any manner of which we have experience, and indeed have 
no need of it. The difference between their vision and ours 
may be almost as great as that between the sense of smell 
in spoi'ting dogs and in a human creature. But it is intelligible 
that in cases where. other peculiarly finely developed senses 
render a more acute vision unnecessary, that there may yet 
exist a morphological visual apparatus which is fairly well 
constructed, but yet is deficient in one of the chemical con- 
stituents which are sensitive to light. It seems preferable to 
ascribe to the bat four senses only, rather than six, as our 
predecessors wished to do. 

Another animal that especially lives in the dark, namely 
the badger, whose eye in relation to the size of its head and 
body must be called small, exhibits a well-developed purple 
in the retina. I received the animal alive, placed it for three 
hours in the dark, and immediately prepared the eye in 
sodium light. The retina bleached quickly in light, passing 

1 Wien, 1875. 



II.] VISUAL PURPLE. 29 

through orange into yellow. I foxxnd the rods very small, 
considerably shorter and narrower than in the rabbit, for 
example. The backgi-ound of the eye shewed a large tapetum, 
not of a triangular but of a half-moon form, in which, somewhat 
corresponding to the centre, near the edge, but in the bright 
part, the entrance of the optic nerve was situated. There was 
no coloured interference to be remarked in this tapetum, only 
a bright satiny sheen. 

Among fishes the eel and the loach, as they are chiefly in- 
habitants of dark mud, may be reckoned as nocturnal animals. 
M. Schultze makes a similar remark in reference to the eel, 
of whose retina he observes that it only contains rods, no cones. 
I have found the retina of both fishes purple : in the loach only 
faintly so, but in the eel more intensely purple-coloured than in 
any other animal, with the exception of the owl, which in this 
respect slightly exceeds it. When exposed to light the retinas 
were often very intensely yellow ; to this fact may probably be 
referred Leydig's statements in reference to the loach; there 
are however differences to be noticed. I saw, for instance, in 
an eel which had died in darkness, the purple pass when 
exposed to the light into a deep orange-yellow which did not 
bleach for two days', and then did not bleach fully, whereas 
in another specimen the yellow came out very faintly and 
disappeared after being exposed for an hour to a dull light. 
As the retina of an eel, being free from cones, contains no 
pigment-globules, and the rods being of a considerable length 
are uncommonly rich in purple, a certain agreement with 
the arrangements in the owl's eye is evident. A specimen 
of Fetromyzon fluviatilis which I examined shewed quite evi- 
dently a faint purple colouring of the retina, which disap- 
peared in the light. The animal came into my possession in 
a condition upon which I could not depend, so that I could not 
be certain that it had not while already dead been exposed 
for some time to the lightl 

It is of the greatest importance to know for certain whether 

1 See Appendix, note B. 

a In another specimen obtained in a perfectly lively state and examined 
•with aU proper precautions the visual purple though present was decidedly faint. 



30 VISUAL PURPLE. [PART 

the purple is always absent from the cones. The sensory epi- 
thelium in man consists in the yellow spot chiefly, and 
in the fovea centralis, in the place of the clearest vision, 
where there certainly is also sensitiveness to colour, entirely 
of cones. Up to this time I have unfortunately only been 
able to examine one pair of human eyes which were in 
any way in a condition in which they could be of any use. 
Dr Fischer, Assistant Physician at the hospital at Pforzheim, 
to whom I am greatly indebted for it, took the precaution to 
shut out the light for about half a minute before death (March 
19th), and to have a dark covering laid over the head and the 
eyes of the corpse. This last came early on the 21st of March 
into the Anatomical School of this place, when the eyes were 
immediately extirpated by Dr E wald under a cloth. The cornea 
was already very cloudy, the bulbs tolerably soft and surrounded 
with a great deal of fat. I opened the first eye by means of 
a circular incision, somewhat in front of the equator. As the 
vitreous humour slid out the greater part of the posterior half 
of the retina came out with it — torn away over a wide extent 
round the papilla. When brought out of the sodium-lit cham- 
ber into the daylight, the posterior surface shewed a very clear, 
uniformly distributed, purple colour, which quickly passed into 
chamois and yellow and finally disappeared. The rest of the 
retina still remaining in the eye was, after an incision round the 
optic nerve, removed under sodium chloride '5 per cent. ; it ex- 
hibited the same colour and sensitiveness to light. The yellow 
spot was very clearly recognizable, there was no red perceptibly 
mixed with its yellow. The fovea centralis was not clearly visible, 
but there was certainly no red spot in the place where it was. In 
the immediate neighbourhood of the yellow spot the retina shewed 
an extremely faint tinge of red, so that the macula was surrounded 
by an almost colourless broad zone, which passed with indistinct 
gradations into the redder parts. The state of the other eye 
was similar ; in this the retina after being halved behind the 
equator was exposed to light quite uninjured, with the exception 
of the small portion cut off round the papilla. There was no 
pigment epithelium clinging to either of the retinas. As the 
eyes were forty- eight hours old, one could not conclude with 



11-] VISUAL PURPLE. 81 

certainty from what was found in them, that the portions of the 
human retina which are rich in cones possess very little purple — 
and that the portions which consist entirely of cones, the macula 
lutea and the fovea, have none, however probahle that may 
be, since the outer limbs of the cones are the most perishable, 
and I cannot be sure that they did not remain partly attached 
to the epithelium, while the corresponding portion of the rods 
remained with the retina. Microscopical investigation could 
not decide this point, for I found the rods and the cones 
entirely changed in the usual way by death. "With reference 
to the colour-blindness of the peripheral portion of the human 
retina, it was interesting to ascertain the limit of the purple 
towards the front. In this I was successful in the second 
eye. I drew the whole contents of the eye out, under salt 
solution, from the sclerotic and the cornea, after which the 
retina was separated with the greatest facility from the uvea 
and the pigment epithelium as far as the ora serrata without 
being torn; from that point to the lens the task was more 
difficult, but I succeeded after careful traction. The last 
separated portion remained covered with brown pigment, which 
allowed no other colour to be visible ; but it certainly did not 
cover up any, for the purple ceased at at least two mm. be- 
hind the periphery of the brown zone, with a boundary line 
by no means diffused. The eyes belonged to the body of 
an old and corpulent woman, whose lenses were yellow, and 
tolerably soft. My observations on the visual purple of the 
human eye confirm, as one sees, those of Fuchs and Wel- 
poner\ Schenk and Zuckerkandl^, but with respect to 
the extent of the purple, touch upon a circumstance of the 
greatest importance, which has not yet been considered, and 
upon which I am anxious to arrive at greater certainty. As 
I had no opportunity of investigating the quite fresh eyes of 
an executed criminal to overcome the doubts into which the 
observation of human eyes two days old had led me', there 
only remained for me to investigate the eyes of apes. By the 
kind assistance of the Director of the Zoological Gardens in 

1 Wiener Med. Wochenshft. d. Js. S. 221. 

" Allgem. Wien. Med. Ztg. 13. Marz, 1877. 

2 See Appendix, note C. 



32 . VISUAL PUEPLE. [PAET 

Hamburg, Dr Bolau, to whom I am peculiarly indebted for 
procuring for me many of the animals used in this work, I 
obtained a good living specimen of Macacus cynortiolgus. The 
ape after being kept in the dark for 24 hours was stupefied 
with chloroform, the head cut off and the eyes immediately 
extracted in sodium light, and further treated as the human' 
eyes were under salt solution. I was not successful in empty- 
ing the vitreous humour well in either eye, or .in separating 
the retina from the choroid coat after the papilla had been cut 
round. For that reason I placed the anterior as well as the pos- 
terior section in alum of 4 per cent, and first drew the retina 
out after 24 hours, an operation which was easily accomplished. 
Both retinas shewed a pale purple colour, with a striking de- 
crease in the part surrounding the yellow spot, in which as well 
as in the fovea there was not the slightest trace of red visible. 
Examined under the microscope there was found in the fovea 
of one of the retinas a very small three-cornered slit, to- 
wards which the cones crowded with their long narrow outer 
limb, and which did not give the impression as if a number 
of cones had remained attached to the epithelium and had 
fallen out, but seemed rather like a rent. In the other eye 
the posterior surface of the corresponding part of the retina was 
continuous, and in all sections which were made from this and 
other parts might be recognized an unbroken border composed 
of the outer limbs of the rods and cones. The alum, it is true, 
caused these structures to shrink, but the rods could be well 
distinguished from the cones, and among the latter could be 
recognized the elongated ones belonging to the fovea. Lastly, 
to make quite certain, the surface of the epithelium of the 
fundus of the eye was scraped off little by little, and examined 
for outer limbs. In the second eye, in which the fovea was 
uninjured, none of these were found ; in the first, isolated 
ones appeared, and here and there several sticking together. 
I maintain, after the results obtained from investigating the 
first retina, that it is proved that in the eyes of apes the fovea 
centralis and its nearest surroundings contain no visual purple, 
while I must leave it undecided, for the periphery of the 
yellow spot, whether the rods which are placed there, and in 



II-] VISUAL PUEPLE. S3 

a wider circle between the cones, possess any purple. In the 
apes the purple extends faintly, as in man, as far as the ora 
serrata. Yet I found the limit of the red somewhat closer to it 
and more diffused than in human eyes. 

When the rods are scattered, or lie between a considerably 
larger number of cones, each of these may be intensely purple- 
coloured, but not recogaizable in the general appearance of 
the surface, if the cones are colourless. Should these latter, in 
spite of our not being able to perceive the purple, still contain 
a trace of this substance, still it must be considered how slight 
this must be if it disappears from observation in the fovea in the 
thick border formed by the extremely long outer limbs. Shades 
of the purple which might correspond to the different degrees 
of sensitiveness to colour in the human retina were, as has 
been already said, not recognizable in this latter, and just as 
slightly in the retina of the ape. Any one who has the oppor- 
tunity of examining a fresh human eye will probably be able 
to decide definitely as to the absence of the purple from the 
spot of the most distinct vision, and to confirm that to which 
my observation on the human eye and the conclusion from 
analogy with the eyes of apes could only give a high degree of 
probability \ 

I was led by the statements of Leydig, M. Schultze and 
others, that in the retinas of snakes no rods are present but 
only cones and these free from coloured globules, to examine 
the eye of Tropidonotus natrix ; and as far as this species is 
concerned I can thoroughly confirm the fact. I have carefully 
searched through a good number of such retinas but I could 
discover no trace of red. Nothing was visible except a yellow 
tiage, so slight as to be scarcely worth mentioning, and which 
moreover withstood exposure to light. Up to this time I have 
found no retina which shews the appearance of the images 
of Leeuwenhoek more perfectlj' than does this; I mention 
this fact as proving by the way that in spite of the smallness 
of this snake's eye one may very well succeed in obtaining a 
normally spread-out retina, and that one can bring the cones 
into view in the direction of their long axes. As they have 
1 See Appendix, note C. 
F. P. 3 



34 VISUAL PUEPLE. [PART 

even in that position no perceptible red colouring, the purple 
must be entirely absent from them. Here again is the retina 
of an animal, which without any doubt can see very well, from 
which the visual purple and indeed all visual colour substances 
are entirely absent. I found the retina of Ooronella laevis ex- 
hibiting exactly the same features. 

I can shew, not as strikingly as in the snake, but never- 
theless quite undoubtedly, the absence of visual purple in 
another reptile. Anguis fragilis has yellow-coloured globules 
(so well known in the lizards), situated at the point of junction 
of the inner and outer limbs of the cones, as well as other 
colourless and similarly refractive structures of this sort. As 
the yellow globules are in many places a considerable distance 
apart, there can be no doubt that there is no visual purple 
between them. I have never been able to perceive in the 
retinas of slow-worms, which have been kept in the dark for 
a long time, anything but a faint yellow colour which is 
unchangeable by light. Therefore I must concur in Boll's 
doubts as regards the occurrence of the purple colour in the 
lizards. 

I found the difference between the purple rods and the 
uncoloured cones very marked in the retina of the carp. 
The retina is of a bluish purple colour, recalling that of the 
eel, but one can see with an unassisted eye that the colour 
is interrupted, and therefore very imperfectly saturated — a con- 
dition of things which is occasioned by the peculiarly large 
number of colourless cones which are scattered between the 
red rods. Boll's statement concerning the purple of the osseous 
fishes is hereby confirmed, and 1 can add the same as to the 
retina of a cartilaginous fish. It is true that the retina from a 
shark, which had been kept in the dark and was brought to me 
48 hours after death, was only a purple-coloured jelly, which 
immediately escaped from the back part of the eye and ex- 
posed a magnificent tapetum which shone like polished silver. 
I could, however, in this mass determine the persistency of the 
colour in the dark, the change to yellow and lastly the com- 
plete bleaching in light. In one of these eyes, which had been 
laid open in the dark for an hour, I saw to my great surprise 



II.] VISUAL PURPLE. 35 

the whole of the retinal mass flooded by a clear purple solu- 
tion, which when poured upon a plate exhibited the same 
behaviour to light as the mass itself. Except the beautifully 
iridescent crystals of the epithelium of the tapetum, which I 
maintain are guanin combined with lime, there were no coloured 
solid particles to be seen in the solution. 

It is well known that the newts have rods with slightly 
conical outer limbs, in which M. Schultze suspected a transi- 
tion, both in form and function, to cones. In a marked manner 
these large structures are always very faintly coloured red. 
It is therefore all the more astonishing to be able to recognize 
very well the faint colouration in these outer limbs, even when 
they are broken off, floating about, and" lying on their sides. 
Perhaps this apparent contradiction — viz. that the colour, 
when looked at in the direction of the long axis appears so 
faint, while when seen in the transverse diameter, of course at 
the lower thicker part only, it is so strongly marked — depends 
on the purple being deposited at the surface only of the conical 
outer limbs, and not, as in other cases, distributed throughout 
their whole substance. 

Incomparably splendid is the appearance of the Salamandra 
maculosa, with its true fairly cylindrical outer limbs of rods, 
the large masses of which render intelligible the more intense 
purple colour of the retina as compared even with that of the 
frog. 

Finally it may be interesting to know that the visual 
purple may appear in intra-uterine life, in rods which have 
never been exposed to the light. I found the retina of a 
foetal calf 65 centimetres in length, which had hair on the 
snout, head, tail and feet, quite clearly coloured purple ; the 
colour when exposed to light, first passed into yellow and then 
entirely disappeared ; in this case the rods were clearly recog- 
nizable under the microscope as delicate short palisades. This 
agrees with Schultze's statement that in sheep's embryos the 
first rudiments of the outer limbs of the rods are already mani- 
fest by the time the hair appears, and also as far as the purple 
is concerned with the observations of Fuchs and Welponer 
(I. c.) on human fcetus of 7—9 months. 

3—2 



36 VISUAL PURPLE. [PART 

In an embryo calf 44 cm. in length, I failed to find any 
rods. The same was also the case in new-born rabbits, in 
which I can confirm Schultze's often misunderstood discovery 
that the outer limbs are scarcely developed. 

In speaking as I have done hitherto of the red colouration of 
the retina in general terms as visual purple, I thereby intend to 
express that it depends on a particular material peculiar to the 
outer limb of the rods, on a substance, that is, composed of 
one or several coloured chemical bodies. This can scarcely be 
doubted, since I have demonstrated how entirely independent 
is the colour of the rods from the numerous structural changes 
which take place in the other retinal structures below the layer 
of the outer limbs. Whatever change obliterates the colour of 
the retina must attack or destroy the colouring substance, that 
is the visual purple itself. In the first part of this treatise I 
have narrated a series of cases in which the retina was handled 
in different ways, rather with a view to shewing that the 
colour is independent of the structure, and that the changes 
induced by light are not the result of changes in the struc- 
ture of the rods, than with a desire of fixing with certainty 
the reactions of the visual purple. The numerous experi- 
ments which have still to be related will afford many illus- 
trations of this truth, so that I need not enter into it more 
fully here. 

Although the facts already known give no cause to suspect 
that the light can give rise to visible changes in the structure 
of the rods^ and of other portions of the retina, I have not left 
unused the opportunity which the observation of the colour of 
so many dark preparations offered to me, in order at the same 
time to study the changes in form of the rods which have been 
so often described with a view of determining what share light 
had in the matter. I can nevertheless only report that the 
well-known opacities and striations, as well as a more or less 
distinct appearance of a satiny sheen, are changes which 
take place in the extirpated retina as rapidly in the dark as 
in the light, and possibly arise from osmotic processes, dif- 
^ See Appendix, note G. 



ir.] VISUAL PURPLE. 37 

ferences of tension, coagulations and the like. These changes 
affect the colour only so far as opaque retinas shew a less 
saturated appearance than transparent ones ; and on this point 
the chief importance must be attached to the fact that this 
change already shews itself at the time when the sheen and 
the refraction in the outer limbs have begun to diminish, 
which events are the first token of that delamination of the 
outer limbs of the rods so carefully investigated by M. Schultze. 
In the perfectly fresh retina it is quite impossible to see any 
striation of the rods corresponding to the laminae, a fact which 
is worth remembering in reference to Zenker's theory of the 
production of standing waves, by means of the rays which 
fall upon and are reflected by the laminae. I do not mean 
that the laminae do not pre-exist, for their appearance as the 
result of so many and various, and easily controlled chemical 
actions forbids the idea that they are merely artificial products; 
but the column of laminae does not behave during life like a set 
of separate laminae as Zenker's theory demands, but like one 
which is composed of glass plates cemented together with balsam, 
while after death it may be compared to a similar series dela- 
rainated by a softening of the cement. The thing which after 
postmortem decomposition becomes a set of laminae seems ac- 
cordingly to behave during life more like a glass rod. The living 
rod is composed not of relatively thick plates with a minimum 
of intermediate substance, but of alternate layers of equal 
thickness, which, though having the same refractive index, differ 
entirely in their chemical composition. M. Schultze 's beauti- 
ful investigations teach us that the visual rods of the cfustaceae 
are composed of conspicuous layers of a purple substance alter- 
nating with a colourless one, the two possessing very different 
imbibition equivalents, that of the coloured ones being dis- 
tinctly the greater. 

Very different from the above indifference of the structures 
of the rods, which only holds good in isolated retinas free 
from epithelium, are the very striking changes of the rods, as 
well as other events not concerned with the purple which 
manifest themselves as the result of the action of light on the 
retina in situ and during life. Since here, however, there 



38 VISUAL PURPLE. [PART 

is at least one important factor at work in the process of 
regeneration, I will at present abstain from entering more fully 
into the matter. 

The reactions which I studied on the retina, and will now 
describe, had for their object the discovery of a solvent for 
the rods or for parts of them, in order to obtain the visual purple 
as an isolated substance. Whatever good reason had formerly 
been given for the existence of this substance, I had from 
the beginning said to myself that such an idea only floated in 
air until the purple had been obtained either in solution or in 
a solid form free from all the remains of its structural sub- 
stratum. For this purpose I could of course only use such 
things as I knew from experience did not destroy the colour 
in the dark; many were thereby at once excluded, such as 
caustic alkalis, concentrated acids, and even the most dilute 
mineral acids, alcohol, &c. 

Former digestion experiments with the retina had shewn to 
myself and A. Ewald' (what has since been fully confirmed 
by Dr Kiihne, through study of digested retina sections in 
the Institute here) that all outer limbs of rods leave something 
behind, possibly an envelope, which on account of its complete 
insolubility in tr3rpsin as well as in peptic acid, and on account 
of its resistance to caustic alkalis, may be considered neuro- 
keratin, as well as a substance which indeed shrinks very much 
during digestion but remains recognizable by its fatty aspect, 
and which is soluble in boiling alcohol and benzol, though not 
in ether or cold alcohol. This last substance agrees with the 
characters (dependent on the presence of cerebrin) of the 
medullary substance of nerves, of which indeed M. Schultze 
supposed the rods to be composed. As is known from 
M. Schultze and Rudneff's researches, the outer limbs are 
rapidly darkened by osmic acid just like the medulla of a 
nerve, yet it must here be remarked that they never assume 
the peculiar steel coloured, almost blue-black tint of the latter. 

If the visual purple is connected with cerebrin I expected 
to be able to free it from that with the help of benzol ; but 
dried or moist retinas of the frog and of the ox thoroughly 
1 Verhandl. d. Naturhist. Med. Ver. in Heidelberg, i. H£t. 5. 



II-] VISUAL PURPLE. 39 

treated, first with ether and afterwards with benzol, never gave 
a coloured filtrate, although the purple was not bleached. So 
also with acetic or salicylic acid or with ether holding ammonia, 
none of the purple passed into solution ; and the same negative 
facts were obtained with etherial fat solutions and with fat con- 
taining benzol. Warming to 50" C. (which did not destroy the 
purple) in pure olive or almond oil, after the retina had been 
previously treated with ether to free it from fat, was also with- 
out any effect; so also digestion with ammonia, glycerine, oil 
of cloves, turpentine, and extraction with chloroform, bi-sulphide 
of carbon, &c. 

Rollett's important observations on the freezing of blood- 
corpuscles have made known a method of separating haemo- 
globin, that is of a colouring matter, from a substratum, which 
on account of ifc containing lecithin and cerebrin may well be 
compared with the medulla of nerves or with the ground sub- 
stance of the rods ; and this seemed to point the way to a 
method of isolating the visual purple, I froze a dozen fresh 
frogs' retinas at — 13° C, thawed them, froze them and thawed 
again four times, and then examined them under the micro- 
scope. The rods were much altered, somewhat thickened, 
doubled in length, and split up into laminae in as beautiful a 
manner as I have ever seen. When I rapidly froze another 
retina directly under the cover-glass of the moist chamber, 
and let it thaw under the microscope, immersing the objective 
in a drop of glycerine, I saw that the change took place 
moderately quickly, and if I slid and pressed the preparation 
the rods stuck together in beautiful transparent red lumps. 
When I, meanwhile, froze my eleven retinas (which in thawing 
had collected a few drops of moisture) together with a few 
drops of J per cent, saline solution, and let them thaw once 
more, I obtained from the whole through a miniature filter a 
colourless filtrate. It may here be remarked in passing, that 
the fluid as well as the lumps of the retina on the filter had 
plainly an alkaline reaction, and that the latter after bleaching 
in light shewed no change in their reaction towards sensitive 
litmus-paper. Since the rods must in this case have given up 
all the fluid they contained, this observation affords a conclusive 



40 VISUAL PURPLE. [PAET 

extension of the former statement (p. 21) as to the unchange- 
ahleness of the alkalinity of the retina through stimulation by- 
light. 

In spite of this first unfavourable result, I still continued to 
make experiments based on the analogy of the chemical con- 
stitution and relations of the red blood-corpuscles, and of the 
medulla of nerves, with those of the red rods ; and I have no 
reason to repent of so doing. Just as bile is a means for the 
solution of the blood-corpuscles, so is it also for the medulla 
of nerves', and even for the fresh axis cylinder. Bile, as is 
known, dissolves lecithin, and indeed it usually contains this 
body. It also, especially on slight warming, as I have satisfied 
myself, dissolves the much less soluble cerebrin. As to the way 
in which it works on the rods of the retina, that once seen will 
never be forgotten. A fresh frog's retina placed in a drop of 
bile under a cover-slip, immediately breaks out into most 
wonderful movements ; at the edge the rods shoot out like 
rockets, and where the bile comes in contact with the separated 
freely moveable outer limbs, these may be seen to curl up like 
worms, and then suddenly, with a jerk, to stretch out straight 
again and shoot forward lengthwise ; it is just at this moment 
that the longitudinal striation becomes first visible, then the 
whole column of the superimposed row of the laminae may be 
seen, and finally the whole vanishes. It often is as if a roll of 
coins was shot out of a tube, or like a cartridge of grape-shot. 
If the drop of bile employed be too small, many rods remain 
unchanged for some time, and one has the opportunity of 
watching the process take place gradually in individual rods. 
This may take place at one end only, or at both ends at the 
same time, or may begin in the middle ; and the breaking up 
at the spot acted upon is preceded by a change in the refractive 
power, difiicult to describe. At times, a somewhat thick canal, 
often possessing swellings, is visible in the axis, around which 
not laminee but rings are disposed, and as these split up in a 
radial direction and thus give rise to a longitudinal striation, 
one can well imagine such a structure of the rods as Hensen^ 

> Compare A. Ewald and W. Kiihne, I.e. 
» Virchow's Archiv, Bd. 39. Taf. xii. Fig. 8. 



II.] VISUAL PURPLE. 41 

represents. It is of course understood that for these observa- 
tions, purified bile, that is to say the watery solution of crystal- 
line colourless ox gall, must be used. I recommend keeping 
the solution under ether, and not to employ cholate prepara- 
tions which have been preserved dry, for it has repeatedly 
happened to me, in this way, to obtain fluids that in spite of 
a suitable alkaline reaction would not properly dissolve either 
blood-corpuscles or the medulla of nerves. 

The most important action of the bile upon the retina con- 
sists in the solution of the visual purple, and the goal of our 
wishes would soon be reached by help of it, if it were not that 
the rods of dead mammalian eyes resist the means. The retinas 
of the still" warm eyes of rabbits and oxen, it is true, very 
readily yield up their purple to the bile ; and when I placed 
in the solvent about thirty ox retinas, only a few hours old, 
removed cleanly under chloride of sodium "5 per cent., I obtained 
a mixture in which, according to all appearance, the purple was 
thoroughly dissolved. On filtration, however, the filtrate came 
through hardly coloured at all, and contained, as indicated by 
the spectrum, a little hEemoglobin which was unchanged by 
light. Corresponding to this one finds in the case of frogs' 
retinas which have been kept moist for 24 hours, that the bile 
does not exhibit the explosive action just described, but only a 
slow breaking up, leading it is true to the same final result. 
During this action it brings about no true solution of the rods, 
but reduces them to a much swollen mass, which, however, 
does not usually stop the filter. I must also remark that even 
the fresh rods are never completely dissolved by the bile, 
however much they may appear to be so ; if the preparation 
is gradually diluted with water, considerable remains, probably 
consisting, in part, of neurokeratin, are always to be met with. 
The cells of the retinal epithelium are very perfectly and very 
suddenly dissolved by bile, both in the fresh and in the dead 
condition ; their dark pigment granules stream away from them 
in all directions, leaving behind them in the case of frogs the 
well-known yellow refractive drops. 

More exact studies on visual purple will only be possible 



42 VISUAL PURPLE. [PART 

■when a new method of obtaining it from dead retinas is found, 
or when one will be able to work with a large number of still 
warm retinas in a yellow chamber in the immediate neighbour- 
hood of a slaughterhouse. The eyes of oxen will certainly 
always be subject to this inconvenience, that some hgemoglobin 
will be dissolved with the visual purple, because the retinal 
vessels invariably contain blood. This it is true has not pre- 
vented my making an experiment with the few ox eyes which 
have been brought to me still warm, and thus obtaining 
some cub. centim. of good clear filtered visual purple solution. 
The experiment, however, was so far a failure, inasmuch as the 
solution did not completely bleach in the light. I could 
not, for example, determine whether the substance giving to 
the retina a distinctly pure red tint contrasting strongly 
with the true purple-red, sometimes even bluish-red, of the 
rabbit and the frog, passed also into solution together with the 
proper visual purple. In all absorption and bleaching experi- 
ments in coloured light, the presence of hsemoglobin would 
prove an insurmountable difficulty. I have therefore for the 
present to content myself with experiments carried out with 
materials at once scanty and difficult to obtain, which I pre- 
pared for myself from the eyes of rabbits and frogs. 

Before I say anything more about these it will be desirable 
to give an account of some experiments which I undertook in 
the meantime with dead ox eyes, to discover other means of 
preparing the visual purple, not because they have already 
led to that end, but because they give further insight into 
the relations of the remarkable substance with which we are 
occupied. 

It is for example possible to obtain from such retinas an 
insoluble residuum which consists only of neurokeratin and 
visual purple. For this purpose the dead retinas of the ox 
eyes are first extracted with purified bile and then with water, 
after this with acetic acid | per cent., which must be removed 
as much as possible on the filter by washing with water. 
Further what remains after the filtration is digested for 24 
hours at 40" C. with an active solution of trypsin, containing 



II-] VISUAL PUEPLE. 43 

1. p. c. salicylic acid', the resulting residue again washed on a 
filter, spread out upon a glass plate, dried in 40° C, and extracted 
with ether and benzol. The benzol solution is then left 
to evaporate, the residue moistened with water and washed out 
with concentrated ammonia, all remains of the latter being 
got rid of by evaporation and washing. All these operations 
must of course be conducted in the dark, using sodium light 
when anything has to be seen. The residue of the retina 
thus obtained is free from fat, lecithin and cerebrin, contains 
no albumin, or nuclein, no mucin, finally no gelatin, since the 
latter is readily dissolved by trypsin after previous treatment in 
a salicylic solution with dilute acetic acid ; it consists entirely 
of the neurokeratin of the retinal nervous epithelial apparatus 
with the purple attached to it. The colour of this mass is a 
deep orange-red, and changes in light into a colourless grey 
within a very short time, direct sunlight bringing about the 
result in less than a minute. Properly dried over sulphuric 
acid it is scarcely changeable in light, but quickly bleaches if 
again moistened. It is seen by this how capable of resistance 
the purple is. In fact this body, whose sensitiveness to light 
apparently exceeds all the hitherto known substances which 
are photo-chemically decomposable, resists attacks which over- 
come most of the constituents of living bodies as well as many 
other substances. 

But more, it is well known that many things which resist 
trypsin digestion are decomposed by putrefaction. During the 
experiments instituted to determine whether the visual purple 
was dissolved by alkaline trypsin solutions (which gave a nega- 
tive result), the various mixtures after five or six hours became, 
as usual, putrid ; nevertheless the purple remained unchanged. 
I have kept the stinking mass of bacteria for more than a 
week, during which a blackish material was deposited in addi- 
tion to the purple ; the latter, however, remained distinctly 
recognizable, and when some of the disgusting mass was poured 
out on a plate, the orange-red was bleached by the light. 

1 Since salicylic acid by itself or in the form of one of its salts bleaches the 
purple when mixed with it, care must be taken to treat the retina with the 
above mixture carefully prepared beforehand. 



44 VISUAL PURPLE. [PAET 

With respect to procuring the materials in these ex- 
periments, for which I was obliged to use about thirty eyes, 
it may here be mentioned for the advantage of future investi- 
gators, that we found no difficulty in the slaughterhouses 
here in being allowed to place black bandages on the eyes 
of the oxen before other preparations for slaughter were made. 
The eyes were removed with as little exposure to light as 
possible, under an opaque cloth, whereupon they were carried 
in a deep covered vessel into the laboratory. "With the assist- 
ance of a trustworthy servant, up to this time no ox eyes have 
come under my hand already bleached. In order to obtain 
the retina with the least loss, I recommend the following pro- 
cedure. The eye is divided along the equator and the vitreous 
humour allowed to fall out with a rush from the posterior half; 
while the eye rests on the table the optic papilla is pressed from 
within with a suitable cork-borer, whereby a circular incision 
is made just round the entrance of the optic nerve, and the 
retina itself can then be easily drawn away under chloride of 
sodium of 05 per cent, with a pair of very delicate hooked forceps. 
In doing this, one must begin at the edge of a portion lying in 
the tapetum, and must guard against tearing it, for if the mem- 
brane once gets into disorder it slips about over the forceps, 
and the rods, rubbed off, flow down as a thick mass. An 
attempt to obtain the rods by shaking the retina in a thin salt 
solution failed on account of the impossibility of filtering the 
mass, which moreover did not settle down clear when allowed 
to stand. 

I have already said that the retinas of frogs and rabbits 
are to be preferred to those of oxen, when pure visual purple 
is required. In the rabbit, the blood-vessels are restricted, 
as is known, to streaks of medullated nerve-fibres, which curve 
away on both sides of the papilla; when the retina is being 
drawn away, after the separation of the optic nerve by means 
of a small cork-borer, both those streaks must be cut away. 
The retina, however, tears so easily that it must always remain 
a most critical business to remove it with tolerable success. 
In the case of the frog, after much loss of time and unnecessary 
trouble with so small an object, which must be the excuse for 



II.] VISUAL PURPLE. 45 

the minuteness of the following account, I have learnt to 
remove the eye and retina in the following manner. The 
animal is decapitated low enough down, the skin of the neck 
taken hold of with a cloth and drawn forwards over the nose, 
the skull divided transversely just behind the eyes, the upper 
jaw cut away on either side and the small remains of the head 
rather removed from the eyes than the eyes from it. I then 
take the eye between my fingers, remove the muscles, and with 
a flat curved thoroughly sharp pair of scissors cut away the 
optic nerve while I compress the eye in such a manner that 
the entrance of nerve bulges out somewhat backwards. In 
doing so one must take care to make the cut as cleanly as pos- 
sible, the resistance of the sclerotic envelope being overcome at 
a single stroke, otherwise the eye will not possess that absolutely 
necessary very small hole without which the retina never 
can be drawn out quite smooth with one pull. Lastly, the 
bulb must be divided and the retina taken out. Retinas to 
which the pigment is attached may be thrown away, as it is 
seldom possible to separate by filtration the finely-divided black 
granules from the solution of the visual purple. It would 
appear that the delicate vascular network which lies on the 
anterior surface of the retina of the frog is removed in the 
manipulations just described, for I have never been able to 
detect any haemoglobin in the solution of the purple thus ob- 
tained. In order to gain this latter I have been accustomed 
for every twenty frogs' retinas to use ^ or at most 1 c. c. of 
cholate solution of about 5 per cent, strength. Each retina 
is placed in this before a fresh one is prepared, and remains 
in it for 24 hours. A very small test-tube containing 
scarcely I'S c. c. served as a vessel, in which the retina is al- 
lowed to sink slowly down without shaking. According to 
all appearance the purple dissolved in the bile possesses very 
slight diffusibility, for even after 24 hours there is seen im- 
mediately above the retina a very intensely coloured zone, 
above which in turn the fluid is quite colourless. By slowly 
inclining the glass to and fro one tries to disperse the colour 
through the fluid, and this on the second day may be done 
repeatedly. Finally the fluid is filtered, the clear upper layers 



46 VISUAL PUEPLE. [PART 

being first brought on the filter by means of a fine pipette, the 
retina being left till last, when all the clear part has gone 
through. It will take about 24 hours until the last drop is 
filtered; if the mass has been shaken, it very often does not 
filter at all. It need hardly be stated that a miniature funnel 
with a very small filter must be used. I have myself taken 
the trouble to prove that all colour may be taken away entirely 
from the residue on the filtering by washing with colourless 
bile ; but I do not recommend this plan of increasing the purple 
by washing, because the slightly coloured wash-water is hardly 
worth collecting, while, on the other hand, the red saturated 
filter paper and the varnish-like homogeneous agglutinated 
retinas when removed from it can be employed for many im- 
portant experiments. 

The filtrated solution of the visual purple is perfectly clear 
and of a beautiful carmine red colour. At first I thought 
I perceived a somewhat blue fluorescence, but I have reason 
to suppose that in the first experiment in which I and others 
fancied we saw this, it was traces of the black epithelium 
pigment, whose appearance deceived us. In the light the 
solution became first orange, then yellow, finally colourless like 
water. The peculiar chamois colour which appears in the 
retina as it bleaches, often occurs in the purple solution but 
not so clearly. In direct sunlight the dissolved purple 
bleaches in a moment; in diffuse daylight with very varying 
rapidity, apparently dependent on the intensity of the light, 
being considerably longer during the afternoon, although our 
eyes at that time seemed to be quite as much, or even more 
affected by the light than in the morning or mid-day. This 
circumstance, which is well known to the photographer and- not 
inexplicable, is also very perceptible in the bleaching of the 
purple while still in situ in the retina. Time and fitting material 
have up to this time failed me for going more fully into the 
chemical reactions of the visual purple^. Whoever considers the 
trouble of preparing the purple will not be surprised that I 
first of all employed it in another direction, and any one who 
remembers the disadvantages which the presence of bile acid in 
^ See Appendix, note H. 



II-] VISUAL PURPLE. 47 

the solution offers to chemical experiments will not wonder that 
I remained content with the results which have been already- 
given until I had succeeded in obtaining the purple dissolved 
in a more convenient medium. Let it at present suffice to 
say that the purple solution, concentrated on ground-glass 
plates at 40° C. or in the desiccator, dries to a beautiful purple- 
coloured varnish which in a perfectly dried condition passes 
even in direct sunlight extremely slowly into orange, and then 
remains unchanged by light until it has been again remoistened. 

It was most important first of all to determine on the 
solution of the visual purple, the absorption of light of different 
wave-lengths, which up to this time had never been accom- 
plished in the retinas themselves by means of the usual methods 
of spectral analysis either with transmitted or reflected light. 
I have repeatedly attempted to become acquainted with the 
absorption spectrum of the retina, by holding the glass plate 
on which it had been spread out before the slit of the spectrum 
apparatus; either the absorption was too weak or there were 
so many horizontal shadows and stripes in the image that 
nothing could be made out clearly. One can never fold or lay 
simple retinas over one another without getting such disturb- 
ances of the spectrum, and I have met with the same difficulty 
when I have brought a pile of frogs' retinas between the slit and 
the source of light. An experiment of leaving the mass to 
swell up in ammonia or trying to make it more homogeneous 
by allowing it to lie in glycerine produced no improvement. 

The solution of the purple brought before the slit of the 
apparatus shewed in all the degrees of concentration prepared 
by me diffuse spectra only ; the absorption begins in the yellow, 
being very weak in the yellow of the D line, increases to E, 
the rise being very rapid at the beginning of the green, and a 
further increase at the junction of the green-blue and blue, and 
decreases again towards the violet. The beginning, transition, 
and end are however so extraordinarily diffuse that it would 
not be possible to determine the curve more accurately without 
a special photometric method. From reasons hereafter to be 
discussed, peculiar importance was laid on the points, whether 



48 VISUAL PtJEPLE. [part 

any absorption was already present at D, and whether the 
diminution in the violet was considerable. I think I may decide 
in the affirmative as regards both points : — for the first, because 
the retina held up before sodium light appears grey when it is 
purple-red, and when more faintly coloured shews grey stripes 
on a clear ground, if the rods have been brushed away in 
streaks; as a further reason may be adduced the indubitable 
weakening of the intensity of the bright D line in the sodium 
spectrum, which is perceived when the slit is for one half 
covered with a glass holding the visual purple, and on the 
other with one holding water. The beginning of the absorption 
in the yellow is also unmistakable when the continuous spec- 
trum of a gas flame is used either with a weak light or with a 
narrow slit, and the same is seen by daylight suitably toned 
down. With respect to the violet, experiments in gaslight at 
first did not leave me in doubt. I here saw the absorption com- 
pared with that in the whole extent of the blue considerably 
less, but when I employed the violet end of the more con- 
venient spectrum of daylight I was doubtful and thought I saw 
the contrary. The contradiction disappears when one reflects 
that the purple in daylight quickly fades into orange or yellow, 
and that was what I saw in the spectrum. If one makes haste 
and employs daylight of just sufficient intensity, one can no 
longer be in doubt that the visual purple solution allows at 
first much of the violet rays to pass through. When this 
is not the case, the eye without any further assistance can 
recognize the beginning of the change into yellow, and which 
corresponding to the spectroscopic image now shews, besides 
a shading of the violet, a marked clearing up from the yellow 
to the green-blue. This behaviour appeared to me not un- 
important, because it renders difficult the easy and in many 
respects seductive assumption that the purple represents a 
mixture of two pre-existing colours, a red yellow and a blue 
violet. If this were so it would be intelligible why in the first 
moments of the illumination so much violet passes through. 
On this point the chemistry of the visual purple must decide. 

Towards the end of the bleaching, it is only the blue and 
violet rays which are held back from the solution, which is con- 



11.] VISUAL PURPLE. 49 

tinually becoming more and more brightly yellow; at the 
conclusion of the decomposition however they again become 
visible. The decolouration is then complete ; the solution looks 
like water. As to what takes place in the ultra-violet rays, for 
reasons to be mentioned below, I wish to defer making any 
statement until I have made further investigations. 

Although the absorption spectrum of the visual purple can- 
not be compared with the well-known spectra, rich in charac- 
teristic lines and bands, afforded by the many pigments and 
coloured solutions which have been so carefully studied, still 
a knowledge of it may furnish many points with reference to 
the relation of the purple to the different coloured bodies which 
are present in the eye or in the retina. Besides nothing justi- 
fies the propriety of its name more than the fact that it princi- 
pally reflects or transmits red, orange and violet, for it is a 
mixture of violet and red which is called purple, and from 
which we produce purple. Lastly, the following will shew that 
the absorption stands in close relation to the chemical decom- 
position of the absorbents in coloured light. 

From the beginning of these investigations I have placed 
before me the aim of making quite clear the behaviour of the 
visual purple in monochromatic light ; and if after three months 
I have not been so successful as I could have wished, the sky 
must bear a portion of the blame, for an unprecedented continued 
clouding of the sun from the first days of January onwards led 
me to seize upon all imaginable artificial methods of compen- 
sating for the loss of the solar spectrum, although only imperfect 
results were to be expected from them. However, as the living 
eye seldom is exposed to monochromatic illumination, I shall 
also give these results below as useful for our vision. 

Obviously the objective spectrum was the means of deter- 
mining the absorptive power of visual purple in its natural 
position in the retina compared with that of its solution — a 
matter which appeared to me indispensable in order to demon- 
strate the identity of the extracted and the pre-existing sub- 
stance. It was indeed chiefly by this means first shewn that 
haemoglobin, for instance, is the natural colouring matter of the 
red corpuscles. I hoped that the dispersion in the retinal 
F. P. 4 



50 VISUAL PUKPLE. [PAET 

membrane which interfered with all experiments in which the 
retina was placed between the source of light and the prism, 
would not prove an obstacle in the case of the objective 
spectrum, a hope which was realized as soon as I placed the 
frog's retina laid on a slip of milky glass in the spectrum of a 
Drummond light. The little moist membrane spread out on its 
front surface and somewhat curled up, presented a convex sur- 
face to the light, and resembled a glistening drop when illumin- 
ated by a weak sodium flame, took the colour of the surface 
on which it rested when exposed to the red, appeared cloudy 
in the yellow and when moved from green to violet looked like 
a glittering black pin's head, while in the violet it seemed dis- 
tinctly brighter and even as if itself of a bright violet sheen ; 
in no part of this spectrum was any appreciable bleaching 
obtained within 25 minutes. 

For experiments with the solar spectrum the following 
arrangements were used. The sunlight, which was reflected 
from the mirror through the slit, fell upon an achromatic Stein- 
he il lens of sufficient aperture, which was placed at twice its 
focal length behind the slit. At half the focal length behind 
the lens an equilateral Steinheil flint-glass prism was placed 
in a position of minimum deviation. For the purpose of receiv- 
ing the spectrum I used slips of milky glass, on to which I 
fastened the retinas in the plane of formation of the images of 
the slit. The retinas readily adhere to the glass, notwithstand- 
ing the vertical position of the support, if one carefully removes 
with blotting paper the fluids of the eye which have come out 
with the retina. Beneath the row of ten or twelve retinas placed 
close together side by side I fastened a strip of paper in order 
to indicate the position of Frauenhofer's lines, and the limits 
of the visible portion of the spectrum. With my arrangement 
the spectrum was 6 cm. long, so that the first retina lay half in 
the ultra-red, the last almost entirely in the ultra-violet. The 
same space could be well covered with two rabbit's retinas cut 
into strips and placed in continuation. When the height of the 
spectrum was 3 cm. one could combine with each experiment a 
second one, by allowing a small section of some chosen colour 
to pass through another slit above the strip of milky glass, and 



n.] VISUAL PURPLE. 51 

after its transmission through a suitably arranged lens refractino- 
it by means of a second prism. I thus obtained, as was done by 
Helmholtz^ a second short spectrum of one colour, whose ends 
yielded on each side a monochromatic light of great purity, 
which could be used for acting on the retina. The intensity is 
of course in this case considerably less than with the first 
spectrum. 

With a width of the first slit ranging from OS to 01 mm., 
which must be altered according to the varying intensity of 
the sunlight as recognized in the spectrum, the action of the 
monochromatic light sufficed for the experiments ; the second 
slit scarcely admits of being taken less than 1-5 to 1 mm. wide. 
I was always careful with the first spectrum to place the strip 
which carried the retinas in such a position that the Frauen- 
hofer lines could be seen distinctly and in large numbers on 
the slip of paper ; they cannot readily be distinguished on the 
milky glass. The D line had before commencing the experi- 
ment been made conspicuous by the use of a sodium flame 
behind the slit, and its position carefully marked. 

Since the retinas of frogs kept in the dark shew strong 
individual differences, one must not omit to satisfy oneself of 
the uniform colour of the whole series of retinas by a hasty 
glance in daylight before they are exposed to the action of the 
spectrum, in order that one may separate out those which are 
too red, or too pale, or too bluish-red, and replace them by 
others; it is also necessary in every experiment to place a 
retina on the strip entirely outside the spectrum, for sub- 
sequent comparison. In addition to this, one must take 
retinas retaining the least remains of epithelium and pig- 
ment, for these make the purple more persistent ; since I must 
here confirm what was expressed as a conjecture in the pre- 
ceding publication, namely, that it is essentially and entirely 
the retinal epithelium which regenerates the visual purple or 
protects that colour from destruction by light. Eetinas which, 
as shewn by the condition of the emptied fundus and intact 
uvea, retain no trace of the choroid proper, but only the pig- 

1 Compare Poggendorff s Ann. Bd. 94 and Physiol. Optik, S. 264. 

4—2 



62 VISUAL PURPLE. [PART 

ment epithelium or its processes, behave in respect to their 
colour as resistant against light as those previously described, 
to which the choroid or pieces of it had been left attached. It 
therefore seems to me probable that one reason why the purple 
of cold-blooded animals is generally longer in bleaching than 
that of warm-blooded animals, is on account of the retinas of 
the former so often retaining functional products of the pur- 
purogenous epithelium. The agents which are concerned in 
the purpurogenous function appear to be soluble or at least 
diffusable', since unmistakable red zones and lines are formed 
during the bleaching in the parts round the shreds of the epi- 
thelium, especially when such a retina is placed in the dark 
after bleaching. Perhaps one may also refer to this the curious 
fact, that in vertically arranged preparations the under surface 
is always the last to bleach, and that when a series of retinas 
are placed one above the other with their edges overlapping, 
the lowest one often bleaches many minutes later than the 
uppermost one, provided that the illumination does not produce 
a too rapid total bleaching. This does not arise from the break- 
ing off and tearing away of entire rods, since it may be observed 
in cases where the surfaces of the rods which are turned towards 
the light are beginning to dry up. As a last precaution in 
starting conclusive experiments with the spectrum, it remains 
to be mentioned that. the retinas must be kept in a uniform 
state of moisture. Drying has such a great influence on the 
time recjuisite for bleaching, especially in that stage when the 
passage from orange to yellow and white occurs, and this influ- 
ence increases so much more quickly towards the red end of 
the spectrum, that this fact alone may lead to the greatest 
deception if it remains unnoticed. In order to avoid this, I 
frequently sprinkle over the preparation with a small spray 
machine, such as is used for freshening up flowers. 

Of the experiments which were made on frog's retinas, the 
details of one of the best, carried out with all the necessary 
precautions by means of the narrowest slit and the purest sun- 
light, may now be advantageously recorded. 

On the 11th of March, at 11 a.m , the sky being cloudless 

' See Appendix, note D. 



II.] VISUAL PUEPLE. 53 

and of a deep blue colour (there was a frost), 10 uniformly- 
purple retinas of Eana temporaria were spread out in a row 
touching each other, and placed in the spectrum in the follow- 
ing manner : 

No. I. a (1st half), in the ultra-red. 

„ I. h (2nd half), „ red. 

„ II. a (1st half), „ orange. 

„ II. h (2nd half), over D to the beginning of the green- 
yellow. 

„ III in the green-yellow and pure green into 

the middle between D and E. 

„ IV. a (1st half), in the blue-green. 

„ IV. h (2nd half), „ green-blue to F. 

" '^- " ^^j!" I between F and G. 

,. VI. „ mdigo ) 

„ VII „ violet. 

„ VIII. a (1st half), „ end of the violet to H. 

„ VIII. h (2nd half), „ beginning of the ultra-violet. 

„ IX. X „ ultra-violet. 

The absorption begins in II. h, ends before Vlil. a. 

At 11.20 a.m., the slit was covered up and the row of retinas 
was examined by the light of a match. III. was already bright 
chamois colour ; II. h, IV., v., VI., VII., were bleached in de- 
creasing series, that is VII. was still most strongly coloured; 
I., VIII., IX., and x., when compared with the test retina, were 
not in the least bleached. 

11.45 a.m., second examination as before. III., IV., V. were 
almost entirely bleached ; VI., VII. very pale, but somewhat more 
coloured than the preceding ones; ii. h was distinctly paler; 
I., VIII., IX. and X., unchanged. 

12.28 p.m., I., VIII. h, IX. and x. were still red ; I. a most 
intensely so ; I. 6 shewed perceptible diminution of the red 
colour, also Vlii. a. 

12.43 p.m., VIII. a was entirely bleached, the outer fourth of 
I. h toward ii. was quite pale, its other portion being chamois 
coloured. 

The experiment was concluded by drying the strip in the 



54 VISUAL PURPLE. [PART 

dark, in order to be able to examine it more accurately by 
daylight, with the least alteration arising from the bleaching 
being continued. The row of retinas then shewed the following 
appearance : l. was reddish-orange, only the edge turned towards 
XI. being yellow ; ii., ill., iv. and V. were very pale yellow ; vi., 

VII. strikingly colourless, and quite white with a bluish tinge 
in the spots where traces of a diffuse black pigment existed ; 

VIII. a was a rosy chamois colour ; VIII. h, reddish chamois ; IX. 
and X. similarly coloured, only more intensely. Even IX. and X. 
and I. can be readily distinguished in colour from that of the 
dried test retina, which still possessed the proper purple tint. 

From the upper portion of the spectrum which stretched 
over the strip of milky glass, a portion of the blue-green a 
millimeter wide was conducted through a second slit on to a 
second combination lens-prism, and two frog's retinas were so 
placed at the second short purely blue-green spectrum in 
such a way that the half of each extended beyond its edge on 
the blue side. The exposure began at 11.5 a.m. ; at 11.25 a.m. 
the halves within the spectrum had become decidedly of a red- 
dish-orange colour, at 11.45 a.m. were distinctly paler but still 
orange, and at 12.30 p.m. were chamois colour. 

After these and other experiments I feel able to affirm that 
the action of monochromatic light on visual purple takes place 
in the following manner : 

(1) Monochromatic light decolourizes and bleaches 
visual purple as white light does, only considerably 
more slowly, corresponding to its lesser intensity. 

(2) Of all kinds of monochromatic lights the fol- 
lowing act with decreasing rapidity: greenish-yellow, 
yellowish-green, green, bluish-green, greenish-blue, 
cyanic blue, indigo blue, violet, later pure yellow and 
orange, still later ultra-violet and red. The extreme 
red and ultra-violet rays are not entirely without 
action, but the commencement of the ultra-violet is 
more active than that of the visible red. 

(3) The transitional stages of visual purple as it 
passes to white, namely the products of bleaching, 



II.] VISUAL PURPLE. 55 

orange, chamois, and pale yellow, are least resistant to 
monochromatic light in indigo and violet, are more so 
at the beginning of ultra-violet than in the range 
from cyanic blue to orange, and are most resistant in 
pure red. 

The effect of spectral illumination might perhaps be more 
striking on the retina of the rabbit than on that of the frog. 
I placed accordingly in the spectrum immediately after the 
above-recorded experiments the two retinas of a coloured rabbit 
which had been removed in salt solution, and cut into strips, 
taking care to place the more intensely-coloured narrow purple 
band which they contain as uniformly as possible in one line. 
The width of the aperture remained 0"1 mm. After six minutes 
(at 12.56 p.m.) the part from D to E was already bleached, at 
1 p.m. it was bleached up to G, at 1.30 p.m. I could still recog- 
nize a purple colouration in the violet and pure red portions, 
somewhat more intense in the violet, here however the mem- 
brane happened to be thicker than elsewhere. At 1.57 p.m., 
bleaching was perceptible in the red and violet and in the be- 
ginning of the ultra-violet. When brought now into daylight 
the latter portion appeared chamois-coloured, the former reddish ; 
even in this case one could recognize as far as the commence- 
ment of the indigo a certainly palish-yellow colouration which 
contrasted distinctly with the colourless portion corresponding 
to the indigo and beginning of the violet. 

The most striking result of these spectrum observations 
(and the one most discordant with my first observations, see 
p. 6, carried out by means of absorption colours, and not by 
monochromatic solar rays) is the most welcome fact that pre- 
cisely those rays which most affect our eyes and therefore 
appear to be the most intense, namely, the greenish-yellow, are 
those by which the visual purple itself is the most changed. 
Having regard to the significance of this fact, I have not omit- 
ted to obtain in the manner above described a second purified 
partial spectrum of this part of the* spectrum; and here un- 
doubtedly the most intense action was visible at that portion 
of a frog's retina which was exposed to the end the least green. 



56 VISUAL PURPLE. [PAET 

The action was here so much greater than in the previously 
described blue-green spectrum, that it could be recognized 
within 16 minutes, and in 54 minutes the purple had become 
changed into the palest yellow. The experiment was conducted 
on the 10th of March, between 12.30 and 1.30 p.m., the sky, 
although cloudless, not being entirely free from haze, somewhat 
whitish-blue in fact, and the less intensity of the sunlight on 
this day, compared with that of the following day (see above), 
manifested itself even in the first spectrum in the much greater 
length of time required for the total bleaching. 

Slight as appears to be the influence especially of the red, 
but also of the violet and ultra-violet rays, it is nevertheless 
certainly not entirely absent, and is probably sujBScient to ex- 
plain the fact that our sense of light can be excited by these 
colours. That portions of the blue and violet, although they 
are a long time in bleaching, in the end do so all the more 
completely, deserves special notice. It would, I think, be un- 
profitable if I were to attempt to carry out any further deduc- 
tions from the facts which have just been established; these 
indeed will readily suggest themselves to every one. I rather 
turn to the less thankful but more necessary task of attempting 
to fill tip the gaps in our knowledge of the facts in the almost 
boundless region thus opened out to us. Want of sunlight 
and of the necessary refractive quartz apparatus, has hitherto 
prevented my taking properly in hand any decisive experiments 
upon the ultra-violet rays, desirable as these would be in refer- 
ence to the observations of Helmholtz and Setschenow on 
the fluorescence of the retina, and especially to the fact proved 
by Helmholtz, that the eye can directly perceive the most re- 
frangible solar rays. In company with Mr A. Ewald, who has 
assisted me in the foregoing investigations in the most active 
and generous manner, I hope to be able soon to communicate 
further results, on this as well as on other points in the 
doctrine of visual purpleS Meanwhile the foregoing experi- 
ments may suffice to shew that in the visual purple there exist 
qualities which one would expect to find in a body serving 
as the means of producing vision by chemical stimulation. 
1 See Appendix, note B. 



II.] VISUAL PURPLE. 67 

Concerning our subsequent observations I will onty here make 
the preliminary statement, that as the year drew on and the 
sunlight became more intense, we found the time of mono- 
chromatic action reduced to about a third of that previously 
given, and that the bleaching scale previously determined 
holds good also for spectra in which the relative width of the 
individual colours was quite different from that which was 
described above. 

I come now to the experiments on impure coloured light. 
As soon as the spectral yellow on both sides of the line D 
proved to be active although slow in its operation, I said to 
myself that the sodium flame could not be altogether without 
effect, although throughout my long use of it I never had 
occasion to complain of it, and the photographers most distinctly 
prefer it to candle-light. Hour after hour and day after day 
have I worked with visual purple and retinas in the sodium 
chamber without experiencing any drawback ; and yet it is 
possible to bleach purple by means of this light alone. One 
has only to bring the retina as near to the flame as the 
radiation of heat will permit (this is best done by means of 
a suitably inclined, mirror or sheet of paper) in order to find 
it bleached to a pale yellow colour. Care must be taken by 
means of a chimney having a lateral aperture to prevent the 
rays from the red-hot platinum wire and sodium nodule falling 
directly on the retina, for it makes generally more than thirty 
minutes difference when the rays which proceed from these 
objects are joined to the pure sodium rays. It is impossible 
to avoid other rays of various refrangibility, which, as is well 
known, are not wanting in the sodium flame. Hence these 
positive results, though they are of use inasmuch as they 
remove the contradiction arising out of my previous negative 
ones, will not enable us to state as certain that it is precisely 
the yellow rays in the sodium flame which are the cause of 
its being active, probable as this may seem on account of their 
great luminosity. I need hardly mention that retinas exposed 
for two hours to the blue flame of a Bunsen burner, in the 
same way as to the sodium flame, shewed no change of 
colour. 



58 VISUAL PURPLE. [PABT 

Thallium and indium when vaporised in a Bunsen burner, 
afford the only means of obtaining real artificial monochromatic 
light ; but I have tried in vain to produce any obvious change in 
retinas by means of these beautiful green and blue flames. 
The light is not intense enough for the purpose, and for 
very obvious reasons cannot be kept up for hours at a time. 
I made an attempt to burn thallium in oxygen, with the result 
that this remarkable metal completely fused with the iron 
spoon in which it was placed. It can be no more burnt in 
oxygen than can zinc. 

My first experiments upon the influence of absorption 
colours upon visual purple gave rise to the suspicion that 
mixed light bleaches the retina the more quickly, the larger 
the number, of the more refrangible rays which it contains. 
This held true for a long series of coloured but not absolutely 
monochromatic lights ; thus the frog's retina bleaches extraordi- 
narily rapidly in the bright blue flame which runs down a 
cylinder filled with nitric oxide containing a few drops of bi- 
sulphide of carbon ; one has only to make the retina follow the 
flame down each cylinder through three or four cylinders in 
succession in order to make the purple completely disappear. 
The same efli'ect is produced by burning bi-sulphide of carbon 
in oxygen. Both flames are well known for their extraordinarily 
rapid action on a mixtxire of chlorine and hydrogen and on 
most photographic preparations. When nitrous oxide is burnt 
after being conducted over cotton wool impregnated with 
bi-sulphide of carbon, the flame thus obtained, which has a 
blue core surrounded by a yellow shell, is far less active. 

A retina placed before the narrowest (1 mm. diameter) part 
of a Geissler nitrogen tube, through which I sent the sparks 
of a small induction machine worked by six Bunsen chromic 
acid elements, bleached in twenty minutes to a bright chamois, 
that is to say, in a time which is short compared with the slight 
intensity of the visual sensation produced by this light, and 
which gives rise to the idea that the effect is largely due to 
the intense ultra-violet rays which are present in it in con- 
siderable quantity. With a fluoride of silicon tube, giving a 
beautiful blue light in which the red rays were very weak, and 



II-] VISUAL PUKPLE. 69 

the blue by far the most intense, and which contained very- 
little violet and so little ultra-violet that it gave rise to no 
distinct fluorescence in quinine, I could not succeed in bleach- 
ing either retinas or solutions of visual purple. Before the 
light of a fluoride of boron tube, on the other hand, the retina 
bleaches in a very short time, namely in ten minutes, to a 
pale violet, and in twenty minutes becomes quite colourless. 
In this light, it is true, all rays with the exception of those 
of the blue are present in considerable quantity, nevertheless 
the violet and ultra-violet rays seem to be the most intense, and 
to give rise to the whitish lavender colour which it possesses. 
When the light was allowed to pass through a layer of am- 
moniacal cupric oxide increased by means of a Hermann's 
Hsemoscope, until all the red end of the spectrum, including 
the green-blue, was absorbed, the retina exposed to it shewed 
no sign of bleaching although a piece of quinine paper held 
in it shewed most distinct fluorescence. The experiment was 
also negative when the copper solution was held before the 
nitrogen tube, but this was probably due to the unavoidable 
loss of intensity undergone by the rays as they passed through 
it. Other experiments on coloured light, obtained by means 
of absorptive solutions, I have made, but these probably will 
not be very instructive to those who live under a more favour- 
able sun than we do. . So far my only means of continuing 
my researches was to expose eyes and retinas in coloured light 
for a longer time than I had hitherto done, and I therefore 
spared no trouble to procure thoroughly transparent colours. 
Coloured glasses could not be depended on, owing to the small 
choice they offered and to the difficulty of obtaining intermediate 
tints. It was absolutely impossible to obtain in this way a violet 
free from blue, and I found that the so-called violet glass might 
much more justly be called purple, for it let through a consider- 
able quantity of red. Red, green and blue-violet could however 
be obtained in sufficient purity in the following manner : red by 
a mixture of carminic acid ' and picric acid (not the so-called picro- 
carmine), blue-violet by a solution of ammoniacal cupric oxide, 

' Compare Eollett (Untersuch. a. d. Inst, zu Graz. Hft. 2, p. 158). 



60 VISUAL PURPLE. [PART 

green by cupric sulphate and picric acid. As a source of light 
I used at first not daylight, which varied to a most ex- 
traordinary degree, but the gas flame of an Argand burner, 
which could be kept at a fairly constant brightness. This was 
enclosed in a lantern of blackened metal constructed with 
radially disposed chambers, in which were placed the vessels 
containing the respective solutions. Very useful for this purpose 
were the angular vessels with the square bases of 10 cm. sq. 
which are used in houses in the ordinary batteries for electric 
bells. The concentration of the solution was graduated by 
directing the spectroscope against the brightest part of the 
flame and then adding material until the spectrum as seen 
through the solution was narrowed to the desired limits. From 
red to orange there was no difficulty ; in the case of green and 
blue, at the edges of which sources of error make their appear- 
ance, such a concentration was chosen that in the case of 
green no yellow was visible on the one side of the spectrum, the 
edge appearing reddish, and no blue on the other side, the edge 
appearing blue-green ; in the case of the blue the concentration 
was such that no green was seen on one side of the spectrum 
(the edge here having often the false appearance of being red), 
while on the other side the somewhat weakened violet could just 
be discerned. In the chamber which was intended to give green 
light, the lantern of course contained two vessels placed one be- 
hind the other, since it would not do to mix the cupric sulphate 
with the picric acid ; on the outer surface all light was cut off 
from the vessels except for a slit of about 6 cm. broad and 10 cm. 
high, before which were placed the retinas or small holders to 
fix a living animal in position. Isolated moist retinas behaved 
with this method very much in the same way as was related in 
my first communication ; except that I obtained a more rapid 
action, since the new arrangement with its more constant source 
of light permitted one to work with a less powerful absorption 
than was the case when I had to guard against the enormous 
variations in the intensity of daylight by always employing 
solutions of great concentration. I strove moreover to quicken 
the effects by washing the retinas in weak saline solution, by 
most carefully removing all portions of epithelium, and by 



n.] VISUAL PURPLE. 61 

throwing the light on the object with a properly inclined mirror 
in order that it might act as powerfully as possible. I thus 
found that in a blue, which seemed to my eye so deep that I 
was unable to manipulate in it much better than in the dark, 
the visual purple bleached in two or three houi-s, while in the 
green, which to my eye appeared very luminous, it took from 
three to four hours. In the red, which produced a most brilliant 
light, the action first became distinct after the lapse of 16 hours, 
and it was not till 24 hours that all traces of colour vanished 
from the retina. An owl's retina with highly-developed purple 
colour required 72 hours before it became perfectly colourless, 
and shewed, even after 48 hours, spots not only of yellowish-red 
and brick-red, but even of real bluish-purple. It will of course 
be understood that this last preparation, as well as all the others, 
was carefully protected from drying during the whole time of 
the long experiment. Several frogs' retinas were, during the 
experiment, cursorily examined in daylight, and some which 
happened to be placed under cover-slips were examined with 
the microscope ; and these enabled me to state that in the blue 
light the stage of orange and yellow colouration was that which 
was most rapidly passed through. Bearing in mind Boll's 
already mentioned interesting discovery of the green rods and 
his remarkable statements about their increase in the eyes of 
living frogs, after these had been kept exposed to coloured, 
that is to green light, attention was especially directed to these 
as well as to the colourless grey opaque rods, which have been 
already described by me. As far as I could see, the matter was 
altogether inconstant; grass-green rods were present even in 
considerable numbers in the preparation which had been ex- 
posed to the three kinds of light. The same was the case also 
with the grey rods; and where the retina before all exposure to 
party-coloured light was rapidly examined in darkened daylight, 
and brought before the lantern still thoroughly well coloured, I 
have not been able to notice any change produced by the co- 
loured light in this direction, certainly no increase of the green 
rods at the expense of the red or grey ones. I have however 
observed that the true grass-green rods keep their colour 
longest in green light, are less preserved in the blue and least 



62 VISUAL PURPLE. [PAET 

of all in the red. A retina taken direct from a dark frog, 
and therefore already provided with grass-green rods, is on 
this account in the most suitable condition for demonstrating 
this wonderful variety, when it has been kept so long in the 
green light that all the red rods have become bleached. Under 
what circumstances the green rods arise I shall discuss later on. 
In passing on now to the behaviour of the purple in the 
living eye, I must first of all recall the regeneration of the 
photographic bacillary layer by means of the retinal epithelium 
which I have previously described, since without this it will be 
impossible to understand what I have to say. Boll' says, "The 
proper colour of the retina is being continually destroyed during 
life by the light which falls into the eye. Diffuse daylight 
makes the purple colour of the retina pale. Longer action of 
direct sunlight (blinding) completely decolourizes the retina. 
In the dark the intense purple colour is speedily re-established." 
All this refers to the behaviour of the purple during life, and 
not to the sensitiveness to light (which it may be added Boll 
did not notice), nor to the behaviour of the retina removed 
from the eye together with the epithelium, which, as we have 
seen, is modified by the processes of regeneration taking place. 
Considering the brevity of the above-cited statement which 
deals with the most important processes taking place during 
life, we ought to give it the widest and most favourable inter- 
pretation ; we may therefore suppose that the author, in dwelling 
upon the regeneration in the dark, having treated both eyes of 
the same frog with an equal and sufficient amount of illumina- 
tion until the bleaching might be supposed to have taken place, 
extirpated one eye, taking care that during the preparation of 
the retina no subsequent bleaching could occur, then brought 
the frog into the dark and after some time determined the 
return of the red colour in the other eye. As far as I can see, 
there is still another but less trustworthy method, namely, to 
carry out the experiment on several frogs, and to examine the 
one set immediately, the other not until they have been kept 
some time in the dark. If these are the experiments on which 
Boll's statements rest, they may be confirmed by slxxj one who 
1 Bericht. d. Berl. Acad. , 26 Nov. 1877. 



II.] VISUAL PURPLE. 63 

cares to repeat them, paying attention meanwhile to the follow- 
ing facts. 

Even the extirpated uninjured frog's eye needs a large 
amount of light in order to manifest even traces of change in 
its visual purple, much more so is the eye still in connection 
with its owner, that is, intra vitam. Without direct sunlight, I 
have never succeeded in a room in obtaining frogs' retinas free 
from visual purple, not even when I allowed the animals to leap 
about before the window the whole day long, whether under a 
glass or uncovered and only secured by a long thread. Only 
when I treated the animal in the same way in the open air (in 
January and February) was I successful ; and even then not 
always, as is intelligible when it is remembered that many frogs 
know how to protect themselves against an injurious light by 
drawing in their eyes and closing their nictitating membrane. 
Against direct sunlight they are accustomed, strange to say, to 
make very little use of this protection. When one considers 
the tremendous difference of intensity between the light of the 
open air and that of a so-called fully lighted room, one would 
expect no other change in living eyes, provided that they possess 
no internal means of protection against bleaching, than that 
which is found actually to occur : in a room nothing but direct 
sunlight will produce a complete bleaching. I recommend 
therefore that such experiments as these should be conducted 
in the open air, placing the frogs under glass, on a white sup- 
port, early in the morning, and examining them about noon. 
The retina will then in many cases be found perfectly colourless, 
without even a trace of yellow, in other cases it will be not 
merely reddish, but a bright purple. It will be remembered 
that in p. 9 it was stated that a removed and bleached retina 
regained its purple after being allowed to remain in contact for 
a few minutes with the choroidal epithelium which in the living 
eye forms its natural support. From this it might be supposed 
that frogs deprived of their visual purple would regain it after 
even a veiy short stay in the dark. It must be remembered 
however that visual purple is, during life, so exceedingly durable 
over the whole retina, that the long and intense exposure which 
is necessary to cause its disappearance may possibly also inter- 



64 VISUAL PURPLE. [PART 

fere with the functions of the regenerative epithelium, and 
indeed a regeneration experiment never produces quite the full 
tone of redness as before, p. 9, and this the less so, the more 
intense the light employed. This certainly is due to changes 
■which the epithelium substance has undergone, as may be 
shewn by the following experiment. The retina is removed 
carefully from the eye and exposed until it has become bleached. 
The fundus of the eye is now exposed to intense light for 20 
to 30 minutes, the retina meanwhile being put back into the 
dark. If the retina is now carefully replaced upon its old 
support, the red colour will scarcely, or not at all, return, even 
though the rods be allowed to remain for hours in contact 
with the epithelium. This might be referred to decomposition 
(death) of the epithelium ; as a matter of fact, however, it is 
not due to this, but to the action of the light, since an equally 
long exposure of the uncovered epithelium in the dark, scarcely 
at all affects its power of regeneration, as may be shewn by the 
positive result of an experiment the reverse of the above in 
which the retina only is exposed to daylight. The epithelial 
regenerator is therefore as well as the retina itself sensitive to 
light ; from which it follows, that there is no need to hurry 
with the preparation in the dark, when it is desired to prove 
the colourlessness of the retina of a frog whose visual purple 
has been destroyed by sunlight. I know no experiment demon- 
strating the regeneration of the purple of the rods by means 
of the epithelial nest in which they lie so simple, certain, and 
independent of manual dexterity, as the following. A frog is 
placed for several hours in the sun, both eyes are then taken 
out in the dark, and one is examined in order to ascertain that 
the retina is perfectly colourless ; this being the case, the other 
eye is kept for an hour or an hour and a half in the dark 
before it is examined ; the retina of this will be found to be of 
a splendid purple colour. The regeneration begins in most 
cases in about 30 minutes, and is generally complete in about 
twice that time. 

Having thus learnt the extraordinary intensity or corre- 
spondingly long duration necessary to bleach the frog's eye 
during life, I could hardly hope to produce changes by means 



II.] VISUAL PURPLE. 6.5 

of weak coloured lights. I made, however, several experiments 

in the open air with my coloured solutions, by placing these in 

the so-called crystallization dishes, the sides of which were 

blackened but the bottom left transparent. I chose, of course, 

those which were sufficiently deep for the purpose. Each of 

these I put on the top of a porcelain capsule in which the frog 

was placed, a ring of black velvet being placed round the lip of 

the capsule, between it and the dish, in order that no light 

should come in at the sides. The general results were as 

follows. The frogs which during the day had been placed 

beneath the blue glasses, the sky being very cloudy, shewed 

the palest purple. In the case of those placed beneath the 

green glasses (the colour being obtained by a mixture of soluble 

Berlin blue and picric acid), the purple was of a redder tinge ; 

it was darkest of all under the red. This agrees very well 

both with my earlier results and especially with the later 

statements of Boll, made before the Accademia d. Lincei. In 

other cases when the sun happened for a while to have broken 

through the clouds, a very sensible decrease of the purple 

was seen even under the red glasses ; sometimes indeed this 

amounted to a complete bleaching, but I cannot be sure that 

in these cases all the light which passed through the solutions, 

and so was able to act upon the retina, really consisted of red 

rays only. As to any distinct relation between the colour of 

the illumination and the prominence of grass-green transparent 

rods, no more was shewn by these experiments than by those 

previously mentioned as conducted on isolated retinas ; that is 

to say, nothing definite came out. All that could be noticed was 

that the grey rods shewed all the more striking contrast colours 

the more intense the colouration of their neighbours ; hence 

they inclined to a blue-green tint in proportion as the retina 

was brick-red, or assumed a pure grey-green tint in proportion 

as the colour of the rest of the rods maintained its true purple 

hue. 

Since these experiments under the open sky gave such 

little result, I could hardly hope to be more successful with 

artificial light, which for other reasons was preferable; still it 

was possible with these to make up in time what was lacking 

F. P. 5 



66 VISUAL PURPLE. [PART 

in intensity. One might well suppose that the short, and, for the 
most part, very dull days of January and February might be 
replaced by a gas-lamp burning day and night, -which, since 
frogs for the most part never go to sleep when it is warm, 
would give their eyes no rest. Many frogs, it is true, took a 
dislike to the lantern and turned their backs to it, and hid 
their heads, or when they were prevented from doing this, shut 
their eyes. In the case of some, however, which were not so 
obstinate, I succeeded very well in finding the purple changed 
after 20 or 30 hours. In the blue light this was carried on 
until there were exceedingly small remains of the purple, which, 
when exposed to daylight, changed into a very pale yellow 
before all colour finally disappeared. After 48 hours indeed, 
the retinas of some of the frogs were found to be totally 
bleached. In green light the loss of colour was very much less, 
and in the red none was seen at all. 

That one cannot by the above-described method expect to 
produce any very considerable change in the retina, will appear 
evident when we consider that the frogs when they simply 
stare, so to speak, through the solutions fail to do exactly that 
thing which is most likely to affect their retinas. In order that 
the greatest effect may be produced, they ought to be con- 
tinually looking about, so that they may expose all parts of 
their retinas to the more powerful rays which come from the 
image of the flame itself, and not simply to the sun-ouuding 
diffused light ; and the movements of their eyes ought to be 
so rapid and carried out in such a way, that the parts of the 
retina exposed to the more feeble light should have no time 
to regain their purple before they are again exposed to the 
stronger and more direct rays. After some positive results in 
the midst of a very large number of negative ones, I was about 
to give up the whole arrangement had I not accidentally made 
the following observation. One of the frogs placed in the blue 
light had kept his eye steadily fixed on the flame, and I found 
in the retina, although it had been exposed to the light for only 
14 hours, a most beautiful image of the gaslight standing out 
perfectly colourless on the deep red ground of the bacillary 
mosaic. The retina came out of the eye wonderfully smooth. 



n.] VISUAL PUEPLE. 67 

without any merit due to me, and lay spread out completely 
flat on the cover-glass, every rod standing upright with its 
pigment-free end uppermost. These circumstances serve to 
render unusually distinct a clear sharply outlined centrally 
disposed colourless spot, such as I found in it; and when I 
examined this with a low power of the microscope, it was 
impossible not to recognize the image of the flame with its 
two tongue-shaped points. This suggested a simple means of 
obtaining at pleasure photographs on the retina. It is only 
necessary to render the frog motionless with urari, and having 
cut away the nictitating membrane, and having caused the eyes 
to project somewhat, by stuffing a ball of paper into the mouth, 
to expose the animal for about two hours before a flame. The 
best distance is about 35 or 40 cm. I can only recommend 
the experiment, however, to those who have plenty of time to 
devote to it, for even the most practised hand will not be able 
to say beforehand whether he will succeed in placing the re- 
moved retina in such a way on the cover-slip, as to render 
distinctly visible any image of the flame which may be present. 
In the various experiments I made with a view of finding out 
the most favourable focal distance, I was however repeatedly 
successful. 



Without any doubt there can be no better proof given for 
the hypothesis that light acts upon the retina in the eye as 
in the camera it does upon the photographic plate, than the 
image of the flame which I have just mentioned. I became 
aware of this in the frog's eye, where it was accidentally dis- 
covered. I had for a long time endeavoured to obtain it in the 
eyes of the larger mammals; with what results my two last 
communications in the Centralhlatt f. d. Med. W. (1877) Nr. 3 
and 4, wiU shew, and when I made the definite statement which 
forms the concluding paragraph of the first part of this work 
(see p. 12), I had already succeeded in obtaining the first opto- 
gram, which with all its imperfection sufficed to make opto- 
graphy credible, and to make one think that the fixation of 
the image in the eye was practicable. 

5—2 



68 VISUAL PURPLE. [PART 

Perhaps I may here be permitted to disclaim any share on 
my part in the popular interpretation which has been given to 
my discovery. Far be it from me to rob of the prize of a first 
discovery those imaginative persons, on this or the other side 
of the ocean, who have seen in the eye of a murdered person 
the image of the murderer. Indeed it is no pleasant thing to 
find a serious study considered as a fit companion for such 
ideas. There are many things which I should like to say on 
this point, which I will rather leave unsaid, and simply express 
the wish, that my readers will expect from me no corrobora- 
tion of the various popular accounts to which my name has 
been in a most unusual manner attached. 

After I had discovered that the changeability of the purple 
of living and dead retinas was dependent only on light, I said to 
myself that it must be possible in the extirpated eye to dis- 
cover after the removal of the object the well-known images 
which the dioptric apparatus throws upon the fundus. The 
method which I adopted for that purpose may not have been 
the best and the shortest, so that the detailed account of my 
experiments, perhaps, will rather shew how one should not pro- 
ceed with work of this kind, than indicate the proper steps for 
arriving at the desired end. I imagine, however, it will not be 
well to omit them in the following account, since in similar 
instances many things have come to light, which even an ex- 
perienced investigator perhaps may not have noticed, because 
they lay in a direction which, though offering great interest, 
was not in his immediate line of work. 

In face of the wonderfully complete knowledge concerning 
the dioptrics of the eye, which, thanks to the conspicuous ability 
of the inquirers of present and past times, we now possess, 
I have little more to say concerning the small inverted retinal 
image, than that I have always attempted to bring it into view 
as sharply as possible before I attempted to fix it on the retina. 
The eyes of albino rabbits answer best for this, as in them the 
light shines beautifully through the sclerotic, and it is well 
known these are frequently used for obtaining the image of 
distant objects. Coloured or even black rabbits are almost 
equally useful, although this might scarcely be believed, from 



II.] VISUAL PURPLE. 69 

the appearance of the dark back-ground of the eye. Neverthe- 
less, the pigment of the retina may be very dark, and yet the 
■whole eye may be fairly transparent, provided the choroid is 
only moderately coloured, and the sclerotic is thin enough, (as 
is the case with rabbits), since the rods reach through the pig- 
ment right to the colourless bases of the epithelial cells. In 
rabbits' eyes the transparency may be explained, when the uvea 
is only slightly coloured, bj'' the retinal pigment existing in a 
very thin layer ; but there are eyes, as, for example, among 
birds, which, in spite of a much greater development of the 
pigment, allow a very good image to be seen through the thin 
sclerotic. In such cases the reason of the transparency is to be 
sought for in the fact, that on the one hand the choroid is only 
slightly pigmented, while on the other the retina, after the 
fashion of a brush made of threads of glass, is really pressed 
through the deep black ground of pigment until it reaches the 
transparent layer immediately below. Retinas which come out 
covered with the pigment-layer, and when looked at sideways 
appear like black satin, will, when the light is allowed to fall 
perpendicularly through them, often create great surprise in 
the minds of those who attempt in this way to recognize that 
projection of the rods through the posterior surface of the epi- 
thelium, to which, as far as I know, attention was first called by 
M. Schultze. The matter seems to me here worth attending to, 
on account of its bearings on the reflection of those rays of 
light whose course is parallel to the long axis of the rods, and 
in reference to the function of the refractive lens-like bodies, 
each of which, lying as it does in front of an outer limb, 
probably forms its partial image at the posterior end of the rod. 
It is also of interest at the present moment, when a great deal 
of discussion is going on, as to whether the visual purple can 
be seen by the ophthalmoscope, from which question the visibi- 
lity of the other pigments must not be excluded. 

I have given up distinct examination of the image in other 
eyes than those of rabbits and birds, since even with the most 
careful avoidance of extraneous light, no image whatever is 
seen from behind ; in dogs, calves, sheep, swine, and oxen, this 
is entirely due to the thickness of the sclerotic, not to the 



70 VISUAL PURPLE. [PART 

retina or choroid pigment, for the image may be seen shining 
extremely well through the latter in the pig's eye when the 
sclerotic has been taken away. 

It is well known that in the living eye of man, in the case 
of blonde individuals, the image can often be seen on the out- 
side of the sclerotic. I do not think however that this is due 
to the slighter development of pigment; on the contrary, I sus- 
pect that in the majority of cases, the reason lies in the tenuity 
of the sclerotic, through which indeed the pigment may be 
seen shining with a bluish tinge. Lastly, in the frog it is 
only at the optic entrance that the light shines through, and 
in the eyes of these animals it is not the sclerotic, which as 
is well known is very transparent, or the retinal epithelium 
which stops the light, but exclusively the strongly pigmented 
choroid. 

Distant objects are seen so distinctly on the back of the 
rabbit's eye, that if one could get equally sharp optograms one 
would have every reason to be content. With objects near at 
hand, such as rows of windows on house-sides, there is this 
disadvantage — that they appear too small; more satisfactory 
results are gained with the images of bright window surfaces 
crossed with dark sashes, such as are formed in an eye placed 
some few feet off them. Images obtained by means of the 
small figured discs, described a little way back, remain 
quite distinct when these are brought as near to the eye as 
30 or even 19 cm. I have attempted to determine this last 
focal distance in many ways, by means of the sodium flame for 
instance, when I was working in the dark. The individual 
features of this came out very nicely, but I have not succeeded 
in obtaining as great perfection as I could wish. I strove to 
avoid all traction and changes of external pressure, by leaving 
the eye in its place in the bisected head, and getting at it by 
taking away the brain, portions of skull and contents of the 
orbit. I failed, however, with every linear object which I used, 
to succeed in what I particularly desired to do, namely, to 
determine the focal distance of the eye by means of Scheiner's 
experiment. The use of a photograph gave me the most 
useful results, the glass negative of a head larger than life-size 



II-] VISUAL PURPLE. 71 

beiQg held before the sky. The outlines and the points of 
correspondence in the retinal image may, as a rule, be very 
well seen when the eye is placed 25 cm. behind the object. 
It seemed useless to attempt anything more exact, since the 
eye is subject after death to progressive internal changes, very 
difficult to follow, which interfere with the distinctness of the 
image. This depends chiefly on the well-known opacity which 
makes its appearance in the refractive media, which is all the 
more inconvenient, inasmuch as it diminishes the intensity of 
the incident rays, particularly those which are optographically 
most active; and, as Helmholtz states, prevents the investiga- 
tion of the dead organ by means of the ophthalmoscope, and so 
renders this instrument useless for determining the refractive 
index of various parts of the eye. Other hindrances are caused 
by changes in the iris, in the accommodation mechanism, and in 
the intra-ocular pressure. The iris immediately after death is 
so narrow that the pupil is dilated to the utmost. Very soon, 
however, this becomes strongly constricted again, and then later 
returns to a medium condition. Accompanying changes in the 
accommodation muscles also probably occur, quite apart from 
variations in the internal pressure. These, of course, from 
their anatomical relations, affect the form and position of the 
lens. Lastly, there are changes taking place in the shape of 
the whole bulb including the cornea. In fact, there occur at 
the same time a large number of processes, all of which must so 
much influence the image, that each eye must be tried by itself. 
As to which is the best method of proceeding, about that 
I have no doubt. One must work with urarised and atropinised 
rabbits, the refraction of whose eye must be determined before 
the optographical experiment by means of the ophthalmo- 
scope. I have not, however, been able to spare time to master 
the necessary manipulation, which, as is well known, requires 
much practice. 

A glance into the opened eye of an animal which is not an 
albino teaches one that after the vitreous humour has been 
removed, or the concave fundus inverted by drawing the 
sclerotic over a convex surface, the visual purple cannot be 
seen with sufficient distinctness to permit a colourless marking 



72 VISUAL PURPLE. [PART 

on it to be recognized even if it existed. By long experience 
and practice I have indeed learnt to tell whether the retina as it 
lies on the pigment will appear red or colourless when it is 
removed, and I can always recognize the visual purple in a 
frog's eye by a reddish-brown shading of the black ground; but 
I could not undertake to distinguish the fainter from the 
deeper portions. The visual purple is most easily recognized 
in situ, when it lies over the slightly-coloured opalescent, or 
rather silver-glistening, tapetum of the dog's eye. It is not 
quite so well seen on the greenish-blue and satiny tapetum of 
herbivora. In neither of these cases, however, is it so distinct 
that one would be able to recognize an optogram on it. In the 
case of the rabbit's eye, which possesses no tapetum, the retina 
must be wholly removed before the image can be seen. This 
is a critical operation, and not unfrequently the tearing of the 
delicate membrane brings a laborious experiment to an un- 
happy end. Unable to overcome these difSculties, I turned 
to albino rabbits, in which, as I already knew, the purple- 
red colour of the retina could be recognized in its natural 
position. 

The rabbit's retina, as far as its colour is concerned, is 
divided in a peculiar manner into two portions, into a pale 
upper and a dark-red lower portion. The optic nerve in these 
animals ascends after its entrance into the orbit high up on 
the eye, being closely applied for a considerable distance to the 
outer hinder surface of the bulb; after piercing the bulb it 
takes a direction at right angles to its former course, and 
diverging on both sides gives rise to the well-known white streak 
of medullated nerve-fibres, which appears to form the only tract 
of the retina bearing blood-vessels. The retina is thus bisected 
by this white streak into an upper smaller and a lower larger 
division. Under the white streak is seen a second, of a dark 
purple-red, of about the same breadth. This is sharply defined 
from the upper lighter-coloured division of the retina, but 
shaded off gradually into the lower redder division, at about the 
level of the retinal horizon, and stretches on either side a long 
way towards the front. The medullated fibres accordingly lie 
entirely in that portion of the retina which is poor in purple ; 



II-] VISUAL PURPLE. 73 

and hence the "blind spot" of the rabhit, which must be 
thought of as a dark broad band, lies in the lower part of the 
field of vision, which we may suppose is but seldom used. The 
redness of this red streak is not occasioned by the blood- 
vessels which lie underneath it, and may always be demon- 
strated in removed rabbits' retinas. Along it the purple is 
far more intense than elsewhere, and it disappears, passing 
meanwhile into yellow, so very much later than in other parts of 
the retina, that it remains for a long time still visible in pre- 
parations whieh have been otherwise largely bleached. Finally, 
of course, it also becomes indistinguishable. I have not in- 
vestigated the point as to what arrangement of the retinal 
elements gives rise to this streak, and can only throw out 
the suggestion, that the slight increase of thickening which is 
observable in it depends on the unusual length of the rods. 
Traces of something similar may also be seen in the eyes of 
many other animals. In the redder lower division of the retina 
of a leucotic rabbit, the purple is seen to be uniformly spread 
over its whole surface, and when the animal is bled to death, 
the remains of blood in the choroidal vessels do not usually 
disturb the beautiful appearance. When the posterior half of 
the bulb is turned over the front half — the lens, which serves as 
an excellent support, being retained in position — the whole 
retina looks exactly like a beautiful rose-leaf If it be then 
brought from the sodium light into daylight, the disappear- 
ance of the colour may be watched until nothing remains but 
that of the blood, which, as far as the veins are visible, is always 
arranged in streaks. With regard to the regeneration and 
relative resistance to light of the purple in such preparations, 
I must repeat, with greater distinctness, what I have already 
said in pages 10, 11. A flap of the retina removed from the 
colourless choroid, and placed by the side of the rest of the 
retina, with its anterior surface towards the light, undoubtedly 
bleaches more quickly than those parts which have been left 
in situ. The difference appears more striking when after 
about half a minute a piece of the latter is rapidly removed 
and looked at side by side with the former isolated flap on the 
same support of white porcelain. Much more striking and 



74 VISUAL PURPLE. [PAET 

visible, even after four times the length of exposure, is the 
result when coloured rabbits' eyes are used ; this I imagine 
arises from the fact, that not only the purple, but also the 
purpurogenous factors of the epithelium are protected by the 
pigment against the action of light. Perhaps I might here be 
allowed to suggest the hypothesis, that the black pigment serves 
as a protection, not so much for the rods as for the purpuro- 
genous epithelium ; for I do not see what harm can be done by 
the light passing through on to the sclerotic, which indeed is an 
unavoidable result of the arrangement of almost all eyes, and 
I understand still less of what use to the rods the laterally 
disposed sheaths of pigment can be, for, as Briicke proved, the 
inner cylindrical surfaces of the rods must reflect all the rays 
which do not fall parallel to their axes. If it be supposed that 
the pigment serves as a means of protecting the outer cylin- 
drical envelopes of the rods from the light which falls between 
them, it is in this case unintelligible why the epithelial pro- 
cesses sometimes contain no pigment, and sometimes have it 
arranged in layers, separated by long intervals. One sees what 
a host of questions further investigations on the epithelium 
will have to deal with. Though it succeeds so easily in the 
frog, I have not yet in the case of the rabbit succeeded in 
restoring the colour to a bleached piece of retina by replacing it 
on the exposed epithelium kept in the dark. Without the 
frogs I therefore should not have been able to bring forward so 
satisfactory a proof. I believe, however, that one may with 
reason suppose that this activity of the epithelium is present in 
the opened eye of coloured rabbits for at least two minutes 
after death, though the energy is continually decreasing. 

My first optographical experiment I conducted in the fol- 
lowing manner : In the completely opaque wooden wall of my 
thoroughly blackened windowless dark chamber, which was 
placed behind another room, serving for the heliostat and for 
optical operations, I bored a hole, which I covered with a circular 
diaphragm of 6 mm. diameter. The optical chamber was closed 
by window-shutters, with the exception of a gi-ound-glass pane, 
upon which the bright noonday light fell. In order to see 
how this bright pane, which was about 5-77 metres from the 



II.] VISUAL PURPLE. 75 

■wooden wall just mentioned, came out as an image in the 
rabbit's eye, I first of all hung over it an intensely coloured 
chrome-yellow tissue paper, and arranged an eye in the follow- 
ing way. An albino rabbit, after being kept 15 min. in the 
dark, was decapitated ; one eye was removed from the head under 
sodium light, somewhat cleared at its posterior surface, and fast- 
ened on to the edge of a cork by means of needles run through 
the remains of the conjunctiva. Thus prepared, the eye was 
placed in position in the dark chamber with the cornea pres- 
sing softly against the diaphragm. The image was visible on 
the sclerotic, on one side of the optic nerve, some length of 
which had been left attached to the eye, and so far beneath the 
point of entrance of the nerve into the bulb, that I was sure 
that it fell on the more deeply coloured division of the retina, 
and could readily mark its place in the appropriate quadrant. 
Thereupon the yellow curtain was removed from the pane, and 
the eye after five minutes' exposure was taken away, divided 
along the equator and examined in feeble gaslight. Since 
I could as yet recognize no image on it, I brought the pre- 
paration out into darkened daylight, and shewed it to several 
witnesses. There was evident on the retina a most distinct 
brighter diffused spot, the small dimension of which corre- 
sponded to those of the image previously seen by me, and the 
position of which made me already sure that it was the optogram. 
Every one of the witnesses recognized the spot as being in the 
same place. The eye was removed from its support, and for 
my own satisfaction I tried to find on the sclerotic from behind 
the previously observed position of the image. In this I was 
completely successful, thanks to the help given by the small 
remains of the ocular muscles, the position of which in refer- 
ence to the position of the image had previously been observed. 
I then thrust a needle through from behind, and the exclama- 
tions of surprise from the bystanders told me at once that 
I had hit the right point, for the needle went straight through 
the pale spot. In a similar manner some thirty experiments 
were carried out, most of them with worse, scarcely any with 
better results, in spite of the most varied changes in the object, 
the focal distance, the time after death, the intensity of the 



76 VISUAL PURPLE. [PART 

light, and the duration of exposure. In no case did we succeed 
in obtaining a regular image : all we got was a mere spot or 
patch. In my notes, out of all this series of experiments, 
I only find one which would justify me in asserting the possi- 
bility of optography. This was the case of an eye in which, 
45 minutes after death, an image of a magnesium flame behind 
red and yellow glasses was thrown upon the eye at a distance of 
30 cm. After 15 sec. exposure to the uncovered flame, there 
was visible on the deep red ground of the retina a white patch, 
the form of which was so like that of the burning magnesium 
tape, and was accompanied moreover by two bright dots, which 
corresponded so well with the pieces of metal continuing to 
burn on the table after they had dropped off, that myself 
excepted every one else was free from doubt about it. 

These first not altogether satisfactory results naturally led us 
to consider and experimentally to test all the possible obstacles 
which might interfere with the desired result ; meanwhile I 
succeeded — accidentally, to tell the truth, and without the inter- 
vention of any human art — in obtaining the most beautiful 
optograms. I had not thought of the simplest method of all. 
I had not remembered that the retinal surface turned towards 
the vitreous humour, which alone I invariably looked at, is not 
the red surface, and that the anterior layers of the retina, be- 
coming opaque in death, would naturally throw a veil over the 
real purple surfaces which would allow one to distinguish very 
well pale from coloured parts, but would not permit one to 
distinguish the outlines and individual features of the images. 
And when the eyes used were so fresh, and the operations 
conducted so rapidly, that the retina might still be considered 
sufficiently transparent to give good results, other difficulties 
interfered with the satisfactory observation of the images- from 
the front — difficulties which arose from the creased condition of 
every fundus everted in the manner previously described, and 
from its moist and glittering surfaces. I have since indeed 
seen in eyes of oxen that the difficulties just mentioned are 
sufficient to render perfectly indistinct all the images occurring 
in the layer of rods. 

Difficult as it is to avoid tearing the retina of a rabbit's 



II-] VISUAL PURPLE. 77 

eye in its fresh and unprepared condition, a very simple artifice 
enables one in the happiest and easiest manner to effect that 
removal and inversion of the retina which has become necessary 
for these optographic purposes. After trying a variety of sub- 
stances which I knew did not injure the visual purple, and yet 
might be trusted as hardening reagents for the retina, I found 
a solution of 4 per cent, potash-alum the most suitable for 
giving the membrane its desired consistency. In using this 
substance I neither desired nor reached any other end ; and I say 
this expressly in order to correct the opinion, which has been 
widely spread, that the alum is used as a kind of photographic 
developer of the image. The eye, opened and thoroughly freed 
from the vitreous humour, was, immediately after receiving the 
image, thrown into the alum solution and allowed to remain in 
it in the dark for 24 hours. My original method of inverting 
the eye and drawing the retina over the anterior half before 
laying it in the solution, I no longer recommend, for I have 
since learned the better plan of punching out the papilla from 
the inside. I therefore now recommend that all the more care 
should be taken to prevent the eye collapsing or becoming 
folded in the alum, for the solution does not make the retina so 
tough but that it readily breaks at any creases which have 
arisen during the hardening. After the action of the alum is 
completed the eye is placed under water upon a leaden support, 
the optic entrance punched out, and the membrane seized at a 
point above the optic disk, that is to say, in the more faintly 
red area on which the images seldom fall. If the operation is 
carried out without a mistake, the retina can often, with a 
slight jerk, be pulled out of the solid sclerotic cup like a deli- 
cate shell, without its collapsing or creasing at all. The inner 
surface now looks perfectly white and opaque, it is only the 
convex posterior surface which is red or rose-coloured, like a rose 
petal. I then bring under the water, upside down, a tiny porce- 
lain capsule hardly bigger than half a rabbit's eye, slip under it 
a suitably bent strip of lead (by means of which I can after- 
wards raise it into a horizontal position) and allow the retina to 
sink slowly with its concave anterior surface upon the convex 
outside of the porcelain capsule. Sometimes it is possible to 



78 VISUAL PUEPLE. [PART 

lift the membrane out of the water, with every part of it lying 
smooth on its little porcelain support. Still it will be as well 
in most cases to snip it a little at the edge (taking care of 
course to avoid the image, the position of which can be rapidly 
determined by the light of a match), and to let it float slowly 
upon its support until it becomes spread out, as free from folds 
as possible. I need hardly say that the optogram is best seen 
when the retina is allowed to swim without any change in its 
curvature. The depth of fluid in which the above operation is 
carried on should not be too great, for it is very remarkable to 
what an extent the purple colour is dissolved out by water — at 
least this is the case in alum preparations. In the ox's eye the 
matter is simple beyond all expectation, for in this, after the 
action of the alum, the retina comes out almost as easily as 
if one had to deal with a leathern sack. The method of ex- 
posing the retina of this animal, when fresh, in saline solution 
has been already described (p. 44). If it is desired to see 
optograms on it, it must be allowed to float. I prefer, however, 
the alum preparations to all others, because in them the red 
and the white are most sharply defined both in the moist and 
in the dry condition, whereas in the fresh ox-retina placed in 
saline solution the optograms look very much as if the colour- 
less parts had been cut out in a pale red etching glass. When 
dried these images are more distinct, although not so good as 
the dried, and far inferior to the wet alum preparations. 

The question of fixing the visual purple is of scientific in- 
terest only so far as the methods which effect it may throw light 
on the chemical characters of this body, or of technical value, 
inasmuch as it is very desirable that, in forming a judgment on 
the effects of the exposure of the retina to light, we should not 
be limited to the short period during which the light is actually 
falling on the eye. Every light that would enable us to see 
anything whatever of the retinal colour will, from the very 
beginning, give rise to changes in the retina, and these must of 
necessity interfere with the exactness of the observation, all the 
more so because they are not supposed to be taking place. 

Here, again, it was an accident which led to the first ap- 
proximation towards a successful issue. I had dried the retinas 



II-] VISUAL PURPLE. 79 

of oxen in the dark on strips of milk-white glass, in order to be 
provided with material for spectral experiments when the sun 
should shine again. In the case of frogs I had observed that 
retinas which were simply dried in the air could be completely 
bleached, especially in sunlight, although the action took place 
more slowly than usual. When at last, after some weeks, the 
wished-for sunlight arrived and remained constant for some time, 
the dried ox-retinas were placed in the spectrum and allowed to 
remain there a long time. To my surprise no diminution what- 
ever of colour could be observed. I laid them in the open 
air in direct sunlight, nevertheless the beautiful orange colour 
remained indestructible. The glass slips were then laid in 
water until the dried retinas had become soft, and these while 
still moist were exposed during the next two days to the sun, 
which, however, only came out at intervals. The colour suffered 
somewhat perhaps in depth, but no real bleaching took place. 
This suggested a method of preserving optograms, by bringing 
the porcelain capsules with the retinas lying upon them into a 
sulphuric acid desiccator, and leaving them there in the dark. 
How long they should be kept in this condition I have not yet 
been able to determine exactly. It is desirable, however, that 
the time should be as long as possible, for the influence of the 
length of the dried condition is most remarkably great. Small 
retinas of frogs and rabbits of course dry very rapidly. In an 
ordinary desiccator, in 24 hours they are as completely dry as it 
is possible to make them by means of sulphuric acid, especially 
if, during the drying, they are lifted up from their porcelain 
supports. In a perfectly dry chamber I have found such retinas 
wholly resist exposure to sunshine even for hours, at least 
nothing could be seen but a certain diminution of the purely 
purple colour, so that an orange tint became more prominent. 
It is sufficient to bring them into ordinary moist air in order to 
render them fairly sensitive to light; they may even become 
completely bleached, just like those retinas which have been 
simply dried in ordinary air. When they have been softened 
with water they bleach at once. When, on the other hand, 
they are kept in the desiccator for several days, for instance a 
week or more, the power of bleaching upon being moistened 



80 VISUAL PURPLE. [PAET 

diminishes more and more, so that one can speak of the visual 
purple in this condition in the same way as one does of so many of 
the colours employed in manufactures, and say that it is " fast." 
I have not yet had a sufficiently long experience to be able to 
say to what extent this durability may be carried. To myself 
and others, however, it is a matter of importance to have dis- 
covered a simple method, only needing patience, which renders 
possible the most leisurely observation of optograms. Any one 
who is unwilling to wait for the action of a long drying can 
make observations by means of the desiccator. 

While still in the moist condition optograms may be made 
durable for about a two-days' exposure by laying the alum pre- 
paration in a weak solution of corrosive sublimate ; in this the 
purple passes, even in the dark, into a very light yellow, which, 
particularly in frogs and oxen, less so in rabbits, resists the 
action of light to a markedly greater extent than the natural 
yellow, which in the normal action of light is the forerunner of 
the completely bleached condition. It is not pleasant, however, 
to see a beautiful purple or orange-red optogram assume such a 
washed-out appearance. 

The changes in their behaviour towards light thus witnessed 
in fresh retinas, or in those which had been dried in the desic- 
cator after treatment with alum, I have also observed, as the 
result of a long-continued drying, in isolated visual purple 
mixed with some amount of the retinal neurokeratin. A filter 
covered with this substance, after it had been dried for a con- 
siderable time, no longer bleached in the sun when moistened 
again ; and it was evident in this case that the indifference 
was due not to the mere fact of the preparation being old, but 
that it had been kept dry, since undried preparations of visual 
purple mixed with keratin — which if kept in the dark under 
pure water remain unchanged for a week at a time without 
undergoing decomposition — bleach very well when exposed to 
light, however long they may have been thus kept. 

Although I had thus attained my immediate object, I 
nevertheless undertook several experiments with a view of 
improving the optograms. In these I attempted to follow the 
steps of the photographers, who use the first traces of the 



II-] VISUAL PURPLE. 81 

products of photo-chemical decomposition in order by means 
of them to accumulate new precipitates on the lights of their 
pictures, either through the reduction of the materials in the 
bath, or through the fixation of precipitates already present. 
This however is a development, not a fixation, and is worth 
attempting to produce on the retina, on account of the interest 
attached to the question whether bleached or genuine visual 
purple possesses reducing properties. Gold, silver, and iron com- 
pounds proved in this respect of no use. The use of osmic acid 
brought to light the interesting fact, that the frog's retina can re- 
main in a 1 p. c. solution of this substance for half-an-hour in the 
dark without destruction of the visual purple. The rods it is 
true become very dark, but their brown is distinctly a reddish- 
brown, which when exposed to light becomes markedly paler, 
and can be very easily distinguished from the colour assumed 
by a retina which had been bleached by exposure to light 
before it was submitted to the action of the osmic acid. Pyro- 
gallic acid gave rise to some amount of browning of the alka- 
line membrane, but there was no destruction of the purple or 
of its sensitiveness to light. Permanganate of potash coloured 
the retina brown, the tint beings of course less deep when 
the solution was mixed with 2 p. c. of acetic acid. In the 
mixture the purple remained constant in a wonderful way for 
several hours, bleaching as usual on exposure to light, with the 
intermediate stages of orange and chamois. My efforts in the 
inverse direction, namely, to find energetic reducing agents, 
which would decolourize the visual purple or prevent it bleach- 
ing on exposure to light, were equally unsuccessful. I made 
use of the mixture of ferrous sulphate or stannous chloride, with 
tartaric acid and an excess of ammonia, employed by Stokes 
for the reduction of haemoglobin, but obtained no change of the 
visual purple and no return of colour in a retina bleached by 
light. I was equally unsuccessful with ammonium sulphide 
or hydric sulphide. After such a behaviour of the purple, no 
surprise could be felt on finding that the changes in it are 
quite independent of oxygen, that for instance it behaves, 
when exposed to light in a stream of pure carbonic acid, exactly 
the same as it does in air, in water, in blood serum. 

F. P. 6 



82 VISUAL PURPLE. [PART 

In ordinary photography, fixing, as is well known, means, in 
the majority of cases, the removal of the excess of the unde- 
composed substances still sensitive to light, and in this sense in 
optography there can be no fixing, because the removal of the 
still sensitive purple in the optogram would simply amount to 
washing out the image. The dry method of fixing the opto- 
gram is something quite different from this, inasmuch as it 
consists in the transformation of the materials still sensitive 
to light into those which are insensitive, the substance itself 
and its colour remaining intact. There is also another differ- 
ence, since the optogram is not a negative but a positive, if, as 
there is some reason for doing, the ground colour of an image 
marked out on the red be spoken of as the dark part. On the 
other hand, the optogram agrees with a direct negative photo- 
gram, inasmuch as the light parts are those which have under- 
gone decomposition and are no longer sensitive, while the red 
parts correspond to those, the decomposition of which, unless 
it were arrested, would obliterate the picture. The choice of a 
method for fixing the image is determined however not by the 
last-named agreement, real as it is, but by the first-named dif- 
ference ; for the object we have in view is the preservation of 
the image, and consequently our efforts are directed towards a 
point which may be neglected in the photogram, but which 
must be secured in the optogram, namely, the bringing the 
material into a stable condition without robbing it of its 
colour. 

The foregoing remarks have, I trust, sufiiciently explained 
the methods necessary for optography, as far as the mechanical 
and chemical manipulations are concerned, to enable any one 
who so desires to obtain results for himself I pass on accord- 
ingly to some of the experiments themselves, in order to afford 
proof of the assertions which I have made, and to keep more 
strictly to the optical and physiological questions. 

The alum method which has been previously described, 
and which is so indispensable in dealing with rabbits' eyes, was 
in the first instance accidentally employed in an experiment, 
which I may well call the ' despair ' experiment. After again 
and again failing to bring out any distinct image on the un- 



II] VISUAL PURPLE. S3 

hardened retina, and when I began to fear that I must have 
exaggerated the importance of the visual purple for the 
act of vision, I determined to conduct the experiment just in 
such a way as it would be conducted by any one who under- 
took the task for the first time, without any preconceived 
ideas. I fastened a living rabbit in such a position in the 
holder that the head was fixed immoveable with one eye 
directed against one of the many windows of the laboratory. 
The bulb was made perfectly fast by means of a thread passed 
through the conjunctiva, the lids were kept open by means 
of a spring holder, in which a piece of black paper with a hole 
about 4 mm. wide, serving as a diaphragm, was placed im- 
mediately before the pupil. The head thus prepared was 
covered for about ten minutes with a black cloth, this was then 
removed, and after two minutes the head was separated from the 
trunk, the eye being closely covered Avith the hand, the ball was 
extirpated and opened in the dark, and immediately placed in 
a 5 p. c. alum solution. On the third day afterwards I had the 
pleasure of recognizing in the removed retina the image of the 
window to which it had been exposed, its arched ends appear- 
ing as white silhouettes on a red ground, and between them 
several smaller clear fields. There was only wanting a distinct 
reproduction of the cross sashes. After this we proceeded to 
further experiments, a short a,ccount of which has been already 
given (Gentrlb. 1. c), but may here be repeated with some 
slight additions. 

The plan of optography in the living eye which I had 
formed before I came to make the experiment just recorded, 
was framed in accordance with the following considerations. 
Since normal vision is evidently only possible so long as a 
balance is constantly maintained between the bleaching of the 
visual purple of the rods on the one hand, and the purpuro- 
genous activity of the retinal epithelium on the other, it is 
obvious that one can only expect to obtain permanent opto- 
grams when that balance is destroyed, either by the epithelium, 
in spite of its continued functional activity, being insufficient 
again to colour the rods, or in consequence of circumstances 
which prevent the epithelium from performing its functions. 

G~2 



84 VISUAL PURPLE. [PAET 

This last cause might be expected to .come into operation in 
the eye some minutes after death, but I failed to obtain opto- 
grams under those circumstances, for reasons which were at 
that time not clear to me. I accordingly fell back upon the 
assumption that the dead eye, and especially the anterior layers 
of the retina, became after death impervious to those rays 
which have the greatest chemical activity. This is probably 
the case, but however does not come into play until an hour or 
an hour and a half after death. This view of the post-mortem 
loss of transparency led me to suppose that the experiment 
must always be conducted on living eyes, but I feared occur- 
rence of regeneration, which, as I well knew, even in the mam- 
malian eye was sufficiently powerful to obliterate the image in 
the short period between the decapitation of the animal and 
the contact of the retina with the alum solution ; this latter 
appearing to be the best means of rapidly killing the epithe- 
lium, while it did not affect the colour. The foregoing experi- 
ment was accordingly repeated on a curarised dog with artificial 
respiration, with the difference that it took place in a room 
with one window, and that I took care to saturate the eye with 
alum after the conclusion of the experiment. 

For this purpose, I had previously connected the carotid of 
the same side with an injection apparatus, in order that I might 
drive a rapid stream of warm alum solution into the head and 
into the eye. The vessels of the neck, which upon the separa- 
tion of the head allowed the injection material to flow back, 
were as quickly as possible clamped. One experiment only was 
carried out on this plan, because a repetition of it was mean- 
while shewn to be unnecessary for my immediate purpose by 
further experiments on rabbits. It will however be worth 
while to repeat it when it is desired to study more fully the 
changes which occur in the equilibrium existing between the 
processes taking place in the rods on the one hand and those 
in the epithelium on the other. 

Meanwhile the already mentioned image of the window 
was obtained, and thus all fears were rendered unnecessary, and 
the doubts which had arisen expelled. The condition of dis- 
turbed equilibrium between the function of the epithelium and 



II.] VISUAL PURPLE. 85 

the bleaching of the purple, which hitherto had been put for- 
ward as a suspicion only, must therefore be considered really to 
exist, and I imagine it will be thought not only not absurd, but 
even altogether natural. We must suppose it to occur in all 
cases where our vision-power is either diminished or annulled 
in consequence of exposure to light, and how easily that may 
occur is known to every one who is versed in after-images, and 
the fixation of sight. I might maintain that we cannot fix our 
sight, without winking, for 30 seconds on a large bright light 
without becoming incapable of seeing it, and I find nothing 
wonderful in the fact that the mammalian eye, after a minute's 
constant exposure to moderate light, is blinded in the spots 
where in the visual image the light fell on the retina, indeed is 
so far blinded that the purple either no longer exists there or 
is visibly decolourized, and has diminished so much, that a con- 
siderable time is necessary before such regeneration can take 
place as to render it again visible. In such cases we must 
suppose the existence of, and indeed on examination shall dis- 
cover the presence of, an optogram that is an after-image in the 
proper sense of the word. 

It may be urged that optography depends on changes in 
the rods only, not in the purple free cones, and therefore on 
processes taking place in a visual apparatus undoubtedly far 
less perfect than that of the cones, or that which exists in the 
region of distinct vision ; and it might further be urged that 
the imperfection of our peripheral vision, in spite of the exist- 
ence of the scattered cones which exist in the anterior region 
of the retina, is in many respects remarkable, and especially so 
because the functional activity of the rods is therein involved. 
Under this view, optographic methods which depend principally 
on an incompleteness of the regeneration processes might well 
be expected to indicate the most varied acts of vision. It is 
well known how rapidly objects seen by indirect vision dis- 
appear when the eye is well fixed ; and this would seem to shew, 
even more strongly than does the persistence of after-images, 
that the regeneration of the purple in the rods is a slow pro- 
cess, and has therefore the same significance as the photo- 
phobia so characteristic of creatures in whose retinas the cones 



86 VISUAL PURPLE. [PAET 

are scanty or absent. When these animals lose their visual 
purple they remain blind until the epithelium has furnished a 
fresh supply, while animals which are provided with cones con- 
tinue, under the- same circumstances, to be able to see by means 
of a second move perfect visual apparatus which, in all proba- 
bility, is the sole organ of specific colour sensations — I mean 
the cones. I have convinced myself that frogs, whose retinas 
had been so completely bleached by exposure to the sun that a 
reddish tinge did not return till they had been kept 30 minutes 
in the dark, could see thoroughly well, and I hope hereafter to 
give a satisfactory account of their ability under these circum- 
stances to distinguish colours. As far as our knowledge goes 
at present, it is extremely improbable that the visual purple 
has any share in colour vision, although of course it must be 
admitted that the visual purple and the rods (apart from the 
cones) enable us not only to see the spectrum, but also to per- 
ceive it as consisting of a series of various tints of grey, very 
much as it appears, in all probability, to the colour blind. Our 
results concerning the occurrence and relations of the visual 
purple are in such close accordance with M. Schultze's hypo- 
thesis of the significance of the cones and rods in reference to 
colour vision, that it is sufficient to refer to it in order to check 
at once all hopes of finding a correspondence between opto- 
graphic results and specific colour perceptions. If, however, we 
succeed, as in ordinary photography, in applying photographic 
developers to the red rods at the moment when the very first 
steps of the decolourisation of the visual purple have begun, 
we must in the end succeed in reproducing optographically any 
thing which we see as dai'k and light in indirect vision, if we 
could disentangle in the optogram the numerous after-images 
which in daily life are so completely woven together. It remains 
indeed as wonderful as before how it is that the inner limbs of 
the rods, by a chemical touch as it were, can feel such minute 
decomposition of the purple ; but that is, after all, nothing more 
wonderful than what is taking place in our olfactory cells every 
day of our lives. 

In spite of regeneration of the purple during life as after 

1 See Appendix, note E. 



II.] VISUAL PURPLE. 87 

death, optograms are possible even during life ; here is the proof. 
After we had obtained the above-mentioned image of the win- 
dow, which, though sufficient of itself to prove the point, hap- 
pened to have been seen by two persons only, a coloured rabbit 
was, on the 16th of January at 11.40 in the morning, placed in 
position, with the head well fixed and the right eye exposed at 
a distance of l^S meters to a square opening in the shutter 23 cm. 
high, 27 cm. broad. The eye was not fixed with threads, for 
it was found that rabbits, after the introduction of the eyelid- 
holder, usually remained perfectly quiet unless a noise happened 
to be made. The opening in the shutter was at the level of 
the lowest row of panes, and covered with ordinary window- 
glass, the rabbit, with its holder, being placed at a somewhat 
lower level on a chair. The direction of the eye was arranged 
as follows : standing with the back of our head against the 
opening in the shutter and looking towards the cornea, we 
shifted the position of the rabbit until the centre of the image 
of the sky on the cornea coincided with that of the cornea it- 
self. No other light entered the chamber. The eye was covered 
for five minutes with a black cloth, then exposed for three 
minutes, whereupon the head was severed, the eyeball extir- 
pated as rapidly as possible by the light of a sodium flame in 
an adjoining dark chamber, opened, and immediately placed in 
a 4 per cent, alum solution. All this was effected so rapidly 
that Dr Ewald, within two minutes after the exposure of the 
living eye was finished, was able to repeat a similar experiment 
on the left eye of the same head. For the proper position of 
the eye in relation to the object one was obliged in this case to 
trust to a lucky hit. 

When on the following morning the milk-white and softened 
retinas were carefully isolated over the whole of their areas, 
separated from the optic nerve and spread out, each of them 
shewed on a beautiful rosy-red ground a sharply defined almost 
square clear image, which in the second eye was quite white, 
and sharply defined as if drawn with a ruler, but in the first 
eye was still of a pale rose, with the outlines less sharply defined. 
The dimensions of the images which in both cases were thrown 
on the red retinal areas amounted to more than a square milli- 



88 VISUAL PUEPLE. [PART 

meter; they of course disappeared in proportion as the ground- 
colour bleached while they were being observed in daylight, still 
their disappearance was slow enough to enable me to shew them 
to several competent witnesses. The first account il. c.) of this 
experiment requires a slight correction, since, through an error 
of measurement, the orifice of illumination was then given as a 
square with a side of 30 cm. ; it was not until the error was 
discerned that the aberration of the optogram from a square 
figure became intelligible, and I remembered distinctly that the 
longer side of the image lay in the direction of the red line of 
partition of which I have spoken before, and the position of 
"which in the retina was horizontal. Later on an isolated rab- 
bit's eye happened to be exposed before the same object under 
quite similar circumstances, and in this an image was visible 
altogether corresponding to the one I have described. 

In order to shew more clearly that the optograms thus 
obtained were really optograms, that is, were true images of 
the objects selected to give rise to them, slight changes were 
made in the latter. The window of my optical chamber seemed 
suitable for the purpose; in order to make the bars more dis- 
tinct I increased their width to about 22 cm. by nailing, boards 
to them. Having replaced the lowest row of panes with 
yellow glass, I fastened an albino rabbit, whose eye was only 
covered with a diaphragm, in such a position in front of and 
below the window, that the distance of the cornea from the 
first colourless pane amounted to 1"75 m. If on-e placed one's 
own eye in the position of the head of the rabbit, one saw 
in looking obliquely upwards nothing but the sky through all 
parts of the window. The distance to the key-stone of the 
arch of the window amounted to more than three meters. 
The animal was covered for a few minutes with a cloth, after 
removing which it was exposed for three minutes to the very 
cloudy and extremely dull sky (11 a.m.) and then decapitated. 
The eye was immediately extirpated and laid open in alum. Two 
minutes later the other eye remaining in the head was treated 
in the same manner. In this second eye upon a rapid look at 
the. everted retina in daylight, no image was visible upon its 
beautifully rose-coloured glistening surface. All the more 



II.] " VISUAL PURPLE. 89 

striking was its appearance after it had laia twenty-four hours 
in alum. The posterior surface of the retina of the eye which 
had been exposed during death exhibited the most complete 
image of the window, with two rows of square panes sur- 
mounted by a semilunar space, all white on a red ground, with 
sharp red crosses. The image began at the red line of par- 
tition of the retina, and shewed in its lower part strong per- 
spective foreshortening, especially of the upper row of panes 
and of the arch. Since the head had been laid in a natural 
position, resting on the lower jaw, the upper part of the 
window would of course be reflected into the lower part of the 
eye, just as it appeared in the image. The strong foreshort- 
ening in the figure is intelligible when the distances of the 
lower and upper window-edge from the eye are borne in mind, 
and an albino eye, which for the sake of comparison was placed 
in the same position in reference to the window, exhibited an 
image of exactly the same character, and so afforded a corrobo- 
ration of the correctness of the optogram. 

The experiment shewed at the same time in a very striking 
manner what excellent optograms may be present in the layer 
of the rods without their being visible when the retina is 
looked at from the front. Far in this albino eye, when the 
retina was simply everted, though the most beautiful purple 
colour could be seen on it, no difference of tint could be re- 
cognized which would enable one to infer the existence of an 
image, certainly at least not of so sharp and large an image as 
really existed there. 

In the case of the first eye, which- had been exposed to 
light during life, the posterior surface of the retina shewed in 
the experiment no regular optogram, but only a hardly per- 
ceptible spotty bleaching, leaving it a matter of doubt whether 
it corresponded in size and arrangement to the image which 
had been formed post mortem in the other eye. The intensity 
of the light of the dull sky remained the same as far as the 
eye could judge in both experiments. The weather indeed 
was as bad as possible, therefore all the more suitable for 
shewing the difference between an intra vitam and post mortem 
action of light. The reason why the optography was unsuc- 



90 VISUAL PUEPLE. ' [PAET 

cessful in the first case must lie in the fact, that during life 
the processes of regeneration are more powerful than after 
death, and the intensity of the illumination and time of 
exposure being such as they were, these processes were able 
to replace the decolourized purple by new material. Since this 
experiment I have made several others in a similar manner on 
living eyes, but I must in this place limit myself to the general 
assertion, that one must use good light if one may hope for 
sharp and colourless optograms, when the exposure for many 
reasons is limited — as when one is experimenting on unnarco- 
tized animals — to a short time, for instance, three to five minutes. 
The further development of optography in the living eye, how- 
ever, must depend on the discovery of methods of following out 
the photo-chemical processes of the act of vision, and therefore 
requires more careful treatment than I can at present devote 
to it. 

Post-mortem optography, on the other hand, as is clear from 
what has gone before, offers no real difiiculty. I myself, un- 
fortunately, had for a long time no idea that the experiment, 
which will probably be classed among the lecture experiments 
of the future, is so simple as it really is. One needs for it only 
a fresh eye and a suitable object placed between it and the sky. 
Daylight is always sufficient, even in the worst of weather. 
Whoever has a skylight at his disposal needs- either no ap- 
paratus whatever, or at most only a dark box, the cover of 
which is formed by the object. - When the head, eye, and 
object have to be directed obliquely towards the sky, a suitable 
arrangement of the most simple character is desirable. 

Of the rooms here at my disposal, I prefer to make use of a 
large chamber of eighteen meters deep, which receives light 
from two opposite sides, viz. by three large windows on each 
side, and from above by two horizontal windows in the middle 
of the roof. The skylight, which serves as an object, is of 
ground-glass and is illuminated by an external glass roof of 
considerable size, having a slight slope and looking to the 
north. The iron frame-work of this window, composed of very 
thin bars, is 316 by 2-7 meters in size, and is at a distance 
of 3'98 meters from an eye placed on the table which is gene- 



II-] VISUAL PUKPLE. 



91 



rally used for the purpose. Parallel with the short side of 
the skylight, boards 60 cm. broad were placed over the panes 
at equal distances from the middle, in order to obtain dis- 
tinctive points in the optogram. In a rabbit's eye, I obtained 
in this manner white tolerably rectangular areas of 6 by 4 mm. 
in size, with two red streaks of 1 to IS mm. in length, crossing 
each other at right angles ; at least this was the case when the 
eye was placed so that the visual axis was directed nearly at 
right angles to the middle of the illuminated space, which when 
the bulb only was used was easily arranged, by placing it in 
the mouth of a test tube, and when the head was used by 
employing a soft folded towel as a support. When the image 
comes out thoroughly sharp in a deep-red retina, there may 
be seen by its side an indication of the second, exterior, 
skylight in the form of a narrow trapezium, perspectively short- 
ened. Often have I even seen on one side traces of three 
streaks of light, reflected by the side windows on to the 
roof. In the eye of an ox I obtained under similar circum- 
stances images of 17 to 18 mm. in length in their largest 
dim-ension, that is to say, three times greater than in the 
rabbit's eye. In order to protect the eye before the exposure 
to light, any covering which comes to hand is sufficient, a box, 
a black cloth, &c. Although it is not absolutely necessary, it is 
advisable in order to obtain quite sharp images, to place the 
eye in the middle of a blackened eard-board cylinder. I gene- 
rally use one about 40 cm. wide and 40 cm. high. 

The large number of experiments which I have made in 
this way might have been expected to have led to an exact 
knowledge of the proper time of exposure and the necessary 
freshness of the eyes employed ; unfortunatelj'', however, I can 
state very little concerning the first point. As far as I can under- 
stand, even an experienced photographer is in this respect no 
better off. He acquires a certain eye as to the state of the sky, 
and regulates the exposure by that, without being able to give 
any precise rules. I have obtained optograms in all sorts of 
weathers, in rain, snow, and hail, and after an exposure varying 
from one to twenty minutes. From eleven to two o'clock has 
always seemed to me the best time of day, the afternoon being 



92 VISUAL PURPLE. [PART 

invariably inferior to the morning. The more intense the light 
and the shorter the necessary exposure, so much the better 
were the optograms. As far as the time after death is con- 
cerned, the following rule proved general : that the exposure 
must first of all be long, then shorter, then after a while 
longer again. If, for example, one eye immediately after de- 
capitation requires five minutes' exposure, the other eye taken 
immediately afterwards will give in three minutes an image 
of equal sharpness on a ground of equally deep purple ; while 
under the same light the eye thirty minutes after decapitation 
requires from five to seven minutes, and in an hour or an hour 
and a half a still longer time to bring out an optogram. I am 
not prepared to say that ej^es which have remained in the head 
an hour or more after decapitation will no longer give satis- 
factory optograms ; indeed, the limit for obtaining a good image 
seems to be in rabbits from about sixty to ninety minutes, while 
the eyes of oxen seem to be useless after one hour. Still I have 
never worked with the latter otherwise than after removal from 
the' head, which I find in the case of rabbits shortens the period 
during which they are available. Most probably this point 
depends largely upon temperature. To save others trouble 
and time, I may as well add, that when the severed head 
manifests snapping movements, and spasms develope in the 
ocular muscles, it is well to destroy the brain completely with a 
gum catheter, in order to prevent the optogram from being 
spoiled. The reason why it is in all cases necessary to make 
the exposure longer immediately after death than when the 
experiment is conducted somewhat later, seems to me to lie 
in the maintenance of the power of regeneration, which ap- 
parently continues the longer the less the eye is touched, and 
therefore lasts longer in the head than in the isolated bulb, and 
is shortest in the opened eye. The necessity on the other 
hand which is felt of increasing the time of exposure again 
at a still later period, I would explain by the assumption that 
in such cases the increasing turbidity of the various media 
of the eye, as well as of the retina, allows chiefly the less 
refractive rays to pass through, which act much more slowly 
on the purple. There seems to be no other explanation pos- 



II-] VISUAL PURPLE. 93 

sible, for when eyes which have been dead some time and no 
longer give an optogram are examined, there may be seen 
on the posterior surface of the sclerotic retinal images hardly 
inferior to those witnessed in quite fresh eyes. That the matter 
is essentially due to the retina becoming cloudy I do not at all 
doubt, for when one uses retinas which are more than an hour 
old, it can readily be seen that they bleach much more rapidly 
on a porcelain plate when they are laid with the bacillary 
layer towards the light, than when the other surface is up- 
permost — a difference which is hardly visible in fresh mamma- 
lian retinas, or in the retinas of frogs. 

Sometimes one obtains optograms which are either im- 
proved or spoiled by a peculiar circumstance. In the alum 
preparation, especially after long exposure, very frequently black 
pigment flakes come away with the retina, and these cannot 
be removed, and so either cover the image to a certain extent, 
or, on the other hand, range themselves so sharply at the sides 
of the figure that the picture seems to be sharpened either 
in a positive or negative sense. This result has several times 
occurred to me in optograms taken during life, and in these 
instances possesses double interest, because it seems to depend 
on those movements of the pigment between the rods which 
was first pointed out by Czerny, and which I can now say 
stand in a definite relation to the exposure to light and to the 
subsequent restoration processes *. The occurrence of this im- 
posed on me the duty of investigating whether optograms could 
not arise, so to speak, in a reverse manner, by the rods being 
torn away during the separation of the retina from the epi- 
thelium and left behind in the eye, where their colour would 
be overlooked in the midst of the dark pigment, while the cor- 
responding portions of the retina would be colourless. On all 
occasions, however, when I have snipped off patches of the 
posterior retinal surface in the white portions of an optogram, 
I have never missed the rods, these remaining most distinctly 
recognizable in alum solution. Everywhere were the rows of 
these structures found continuous, in the white parts quite as 
much as in the red. 

* See Appendix, Note F. 



94 VISUAL PURPLE. [PART IL 

Far more striking and more delicate pictures than those 
already mentioned I have obtained when very near objects 
were brought before the eye by means of the pasteboard 
cylinder just described. This consisted of two tubes sliding one 
within the other, so that the distance between the eye and the 
cover which bore the objects, and which consisted of a ground- 
glass table, or an oil-paper stretched on a frame, could be 
varied from ]8 to 30 cm. With this very primitive apparatus 
I have procured a large number of ojptograms, generally under 
the open sky. Objects were formed on the transparent cover 
by means of strips of black opaque paper, about 4 cm. broad, 
and placed at the same distance from each other. Over these 
strips were placed card-board discs, cut out in the centre into 
a square, triangular or circular figure, so that by varying the 
position of the angular diaphragm thus formed to the direc- 
tion of the stripes, or by turning the whole object roiind the 
axis of the animal's head, the most varied pictures could be 
formed. It would be useless to describe all the optograms 
so obtained, but they all agreed in this, that they afforded 
diminished copies, such as might be expected, of simple geome- 
tric figures. When the eyes were placed in an oblique position 
the optograms were correspondingly distorted. 

It is not the duty of physiological optics to bring opto- 
graphy to such a perfection as it might acquire in the skilled 
hands of a professional photographer. I could not, however, 
deny myself the pleasure of treating optographically a few 
complicated objects, such as the garden side of the laboratory 
here and the portrait of a man. Both as yet leave much to be 
desired. In the case of the house the row of windows was 
unmistakeable, in the head (the object was the very large glass 
photograph spoken of in page 70) one could recognize only the 
outline, the limits of the hair, the beard and the shirt collar. 
Whoever took the trouble would probably be still more suc- 
cessful, and by means of such objects might approximately 
determine to what limits the photo-chemical destruction of the 
visual purple runs parallel to difference in the intensity of light. 



APPENDIX. 



Since the foregoing text was published, and during the time the 
translation has been in preparation, investigations have been earnestly 
carried on by Prof. Kuhne and his school, the results of which 
have been published in the Untersuchungen aus dem Physiologischen 
Institute der Universitat Heidelberg. Heidelberg, Carl. Winter, 
1877 — 1878. To translate the whole of these would have made 
this little treatise far too bulky. I have therefore added, in the 
shape of notes, brief accounts of some of the most striking results 
of these inquiries. They will serve it is hoped to fill up gaps in 
the text, as well as to indicate the directions in which progress 
continues to be made; in some instances, it will be seen that 
further inquiry has not corroborated the impressions first made ; 
but in many other instances, the later experimental confirmation 
of early guesses is very happy. 

M. F. 



Note A. p. 5 *. 
Chamois. 

This colour, the name of which, derived from its being one of 
the tints visible in the skin of the chamois, is probably unfamiliar 
to the English reader, deserves some little attention. Like purple 
it is not a spectral colour, but it has the same claims as purple 
to be recognized as a distinct sensation. 

Chamois is a mixture of violet (or purple) with yellow; and 
various shades of chamois may be obtained by varying the admix- 
ture of white light with the above two colours. Violet and pure 

* Untersuch. Bd. i. Hft. 4, p. 463. 



96 VISUAL PUEPLE. [APP. 

yellow (that is, yellow free from green) are not complementary ; tlieir 
mixture therefore does not give rise to white ; but the resulting 
sensatioQ is, as in the case of mixing red and violet, something 
sui generis ; it contains no shade of green or blue, however much 
either of the two component colours, the violet or the yellow, may 
preponderate ; it is simply chamois, just as purple is simply purple. 

It may perhaps be best seen when pure spectral violet is 
looked at in the light of a sodium flame, if a varying amount of 
white light be admitted at the same time as the various shades of 
chamois are obtained. Purple obtained by mixing the two ends 
of the spectrvim also becomes chamois when further mixed with 
white light. This result happens because the pure red of the 
spectrum diluted with white light becomes yellow, while the violet 
similarly diluted becomes lilac, and lilac mixed with yellow forma 
a light chamois. Such a conversion of purple into chamois by 
mixture with white light takes place the more readily the more 
yellow the tint of the red employed. If the violet preponderates 
over the red in the purple, the result of the admixture of white light 
becomes more and more a lilac ; Helmholtz * speaks of the mixture 
of yellow and violet as a whitish rose, but the exact tint of the 
rose colour thus produced will depend on the amount or character 
of the red present at the same time and vary from lilac to chamois. 

Mixed with grey, chamois becomes a distinctive brown, in which 
a violet tint may be readily recognized. 

Chamois occurs in various shades in many tea roses ; many 
flesh-tints are also in reality chamois, and the tints known as 
"Isabel Yellow," "ISTankeen" and "Havannah" are so many various 
shades of chamois. 



Note B. pp. 6, 11, 56. 
Visual Yellow and Visual White. 

The view expressed on p. 18, that visual purple is not the 
only photo-chemical substance present in the retina has already been 
realized. Kiihne and Ewaldt have shewn reasons for thinking 
that the visual purple in being bleached becomes converted into 

* Physiol. Optik p. 279. 

f " Untersuchungen uber den Sehpurpur." Untersuch. a. d. Physiol. Inst, 
d. TJniversitdt Heidelberg. Bd. i. Hft. 1 — 4. 



APP.] 



VISUAL PURPLE. 97 



a substance which they have called " visual yellow," which in tui-n 
by photo-chemical action is further changed into "visual white." 

Visual purple during bleaching becomes chamois and then orange 
and yellow before finally becoming colourless. Very frequently 
(p. 11) attention is arrested by the distinctly orange or yellow 
colour of the retina. Now true unbleached visual purple contains 
no yellow; the red factor forming its hue has little or no yellow 
element. Hence when pure visual purple is diluted, it becomes 
rose-coloured or lilac, growing fainter and fainter, the greater the 
dilution, without ever shewing any tinge of yellow. By dilution 
it does not become chamois or orange. It is only when photo- 
chemical action intervenes that the yellow tint is visible. And 
this is true both of the visual purple in solution and of the natural 
visual purple as existing in the rods. 

From this it is inferred that the passage (during the action of 
light) of the purple through chamois to orange and yellow indicates 
that the optical effects of the disappearance of the purple are 
modified by the appearance of a second yellow colour, a product 
of the decomposition of the purple. And this view is borne out 
by the changes in the absorptive power of the purple as it is be- 
coming bleached. As is stated in the text (p, 47) with unchanged 
normal visual purple, absorption commences very faintly at 
Frauenhofer's line C, begins to increase very rapidly beyond D, 
reaching its maximum between D and E, and then declines, though 
the diminution hardly becomes marked till beyond F, with a very 
rapid decrease beyond G. "When a solution of visual purple is 
brought before the spectroscope at the stage where by the action 
of light it has become a pure yellow, the absorption is very different. 
There is no absorption at all until about midway between D and E, 
and no marked increase until about midway between b and F, just 
where the absorption of the purple itself is beginning to diminish. 
It increases very rapidly at F and reaches its maximum a little 
before G, from which point to the end of the visible spectrum 
the decrease is very slight. The absorption-features of the two 
things therefore, visual purple and visual yellow, are, corresponding 
to their hues, entirely different. 

The evidence of the existence of a " visual white " is as follows. 

It has been shewn by Helmholtz and others that the retina 
is fluorescent : examined in the ultra-violet rays it gives a blueish 
fluorescence. According to Kiihne and Ewald this fluorescence 



y. P. 



98 VISUAL PURPLE. [APP. 

belongs to all the layers of the retina and is visible in the retina 
under all circumstances. When however a retina has been com- 
pletely bleached, brought to a thoroughly colourless condition, the 
fluorescence is not only increased but diflfers in character from that 
of the unbleached retina. In the ultra-violet rays the fluorescence 
is not blueish but greenish. The cause of this new fluorescence 
is confined to the bacillary layer, in fact to the outer limbs of 
the rods, for if these be brushed oflT, the greenish fluorescence dis- 
appears, and isolated bleached outer rods when seen together in 
masses exhibit the same greenish fluorescence. 

Thus there seem to exist three visual colour substances ; visual 
purple or Rhodopsin, visual yellow or Xanthopsin and visual 
white or Leucopsin, the first passing by the action of light into 
the third through the intermediate stage of the second. Visual 
yellow is very apt under certain circumstances, as when the retina 
is dried, to become indolent and to resist the further action of 
light. The orange retina of the rabbit, mentioned at p. 11, was 
evidently one in which the visual purple had been largely changed 
into visual yellow. 

If under the action of light, by the conversion of visual purple 
into visual yellow, the absorptive power of the retina (or of the 
solution of visual purple) is thus itself changed, one might fairly 
expect that photo-chemical action would also vai-y in correspondence 
to the changes in absorption. This Kiihne and Ewald have found 
to be the case. 

It has been stated, p. 54, that the rays most efi'ective in bleaching 
the visual purple are the greenish-yellow and yellowish-green, i.e. 
precisely those rays which are absorbed with the greatest readiness 
by the visual purple. But this is only true so far as the con- 
version of visual purple into visual yellow is concerned. 
Kiihne and Ewald have shewn that the conversion of visual 
yellow into visual white is most readily efiected not by the 
yellow-green but by the more refrangible rays, in other words by 
those rays which are most readily absorbed by the visual yellow 
itself The less refrangible yellow-green rays, while they rapidly 
convert the visual purple into visual yellow, act very slowly in 
carrying out the further change into visual white, while the more 
refrangible blue rays, though slow to convert visual purple into visual 
yellow, rapidly change the already formed visual yellow into visual 



99 



■*-Pl'-] VISUAL PURPLE. 

white. Thus under blue light the retina or purple solution becomes 
colourless more easily than under yellow-green light, while the 
initial partial bleaching into chamois or yellow is most rapid under 
the latter kind of light. Kiihne and Ewald have formulated their 
results as follows : — 

1. All visible light decomposes visual purple; but, intensity 
remaining the same, in very different times, viz. in times proportional 
to the absorption of monochromatic light. 

2. Rays of such wave-lengths that they most rapidly convert 
visual purple into visual yellow act most slowly on the latter : those 
which most easily change the visual yellow into visual white, and 
which are most readily absorbed by the visual yellow have, as a rule, 
a weaker action on the visual purple itself. 

These separate actions on visual yellow and visual white explain 
the difficulties met with in determining thei bleaching power of the 
different rays, before the distinction was appreciated. 



Note C. pp. 31, 33. 
Visual Purple in the Human Eye. 

Since the paragraph in the text was written Prof. Kiihne has 
had opportunities of extending and corroborating the observations 
there recorded*. He finds that the purple is present in the rods 
only, being entirely absent from the cones. To the naked eye the 
fovea and even the macula appear colourless, while in a zone of 
about 2 mm. round the macula the colour is extremely faint. Under 
the microscope, however, the rods in the outer margin of the macula 
are seen to possess purple. 

At the peripheral margin of the retina, a zone of about 2 mm. 
in width lying immediately behind the ova serrata is completely 
colourless, and here the want of colour is due to its absence from 
the rods, and not merely to the scantiness or absence of the rods 
themselves. 

Kiihne's observations have, in the main, been corroborated by 

* " Ueber die Verbreitung des Sehpurpnrs im menscUichen Auge. Weitere 
Beobaehtungen iiber den Sehpurpur des Menachen." Untersiich. Bd. i. Hft. 2. 
" Nachtriige zu den Abhandlungeu iiber Sehpurpur." Ibid. Hft. 4. 



100 ■ VISUAL PURPLE. [APP. 

other observers, including Bonders*. In many cases, it is true, the 
attempt to discover visual purple, even in eyes fi-ee from disease, 
examined apparently with all proper precautions, has failed t; and it is 
of course possible that we have not at present mastered all the conditions 
determiuing the presence or absence of visual purple under various cir- 
cumstances; but on the whole it seems more probable that these excep- 
tions were after all due to defective manipulation. Kiihne points out 
a possible source of error in the rods being so closely attached to the 
pigment epithelium that their outer limbs remain in the fundus when 
the retina is removed ; in such a case of course the retina would 
possess no purple. 

It is more than probable that individual variations in the in- 
tensity and even in the distribution of the purple may be common in 
man; but Ktlhne denies that the intensity of the colour itself, allow- 
ance being made for the less saturation due to the abundance of cones 
in the human retina, is less in man than in the rabbit. 

Note D. p. 52. 
Auto-regeneration of the Visual Purple. 

Kiihne and EwaldJ have shewn that a bleached frog's retina 
may regain its purple even in the absence of the retinal epithelium, 
by a process which may be spoken of as " auto-regeneration." If a 
retina bleached until it has become quite colourless be kept in the 
dark for some time it will to a lai-ge extent regain its purple, 
becoming first yellow, then chamois, and finally of a distinct rose 
or lilac colour. This colour will disappear when exposed to light, 
and reappear once more in the dark. In fact the bleaching and 
regeneration may be repeated several times, the revived colour be- 
coming of course fainter on each occasion. The same "auto-regenera- 
tion" may be observed in solutions of visual purple prepared with 
solutions of bile salts free from ether. These facts suggest the 
idea that the visual purjsle may be formed out of some substance 
in the retina, by an action analogous to that of a ferment, and 
the view at once readily occurs that the body which is thus 
converted by a ferment action into visual purple is in reality 
the visual white. This view is supported by the facts (1) that 

* Beilageheft d. Klin. Monatsbl. f. Augenheilk. xv. Jlirg. 5, 156. 
+ Cf. Michel. Centralbt. f. Med. Wiss. 1877, p. 433. 
± Untersuch. Bd. i. Hft. 2. 



APP-] VISUAL PURPLE. 101 

the isolated retina becoming fluorescent (see note B, p. 97) as it 
bleaches, loses its fluorescence as it becomes once more purple : that 
is to say, the return of the visual purple is accompanied by the dis- 
appearance of the visual white ; (2) that the retina bleached in situ 
during life is found when removed to be incapable of auto-regene- 
ration, and is at the same time devoid of fluorescence, — that is 
to say, when the retina is bleached in situ the visual white appears 
to be removed from the rods as it is formed. It is no great stretch 
to suppose that it is taken up by the retinal epithelium, and there, in 
the epithelium cells, reconverted into visual purple. And this further 
view is again supported by the facts that a solution of the retinal 
epithelium in bUe-salt solution, free from ether, is rose-coloured in 
the dark, becomes colourless when exposed to light, and rose-coloured 
once more when returned to the dark. The best results, however, 
are obtained when a solution of the epithelium and the rods mixed 
together is employed. 

These facts would seem to indicate the existence of a continual 
intercourse between the epithelium and the rods, the former absorbing 
from the latter the visual white arising from the action of the light 
on the visual purple, and by the action of a ferment body (to which 
the provisional name of Rhodophyllin might be given), reconverting 
it into visual purple, and returning it to the rods ; in the isolated 
retina, however, one must suppose the Rhodophyllin to be present in 
the rods themselves. Further, one might readily imagine that such a 
secretory activity of the retinal epithelium would be placed under 
nervous guidance. But the whole of this part of the subject needs to 
be elucidated by further researches. 

Note E. p. 86. 
Vision without Visual Purple. 

The facts that visual purple is never present in the cones, that it 
is absent from the macula lutea and fovea centralis, and that some 
animals (and these not all nocturnal, ex. gr. the pigeon and the fowl) 
possess no visual purple at all, prove very conclusively that the pre- 
sence of this substance is not essential to even the most accurate vision. 
It is impossible, however, to suppose that the changes of the visual 
purple have no relation whatever to vision; and the interesting 
question arises, what differences in respect to vision can be detected 
in animals when their visual purple has been completely bleached as 



102 VISUAL PURPLE. [APP. 

compared with the same animals when their visual purple is intact 1 
Prof. Kuhne* has attempted to answer this question as regards 
frogs ; and though the results are negative, they are worthy of attention. 
The retinas of living frogs may be completely bleached by direct 
sunlight (in summer) in about fifteen minutes, the colour returning 
in about an hour, the first traces being visible in about thirty minutes. 
Frogs, therefore, may be obtained which for about half an hour 
possess absolutely no visual purple at all. Kiihne satisfied 
himself that during this half hour the frogs were able, as far as could 
be ascertained, to see as usual. They fled like ordinary frogs when a 
hand was stretched out to seize them. They caught flies with the 
nimbleness of an ordinary frog. Like ordinary frogs they chose the 
shade, overcoming all manner of difiiculties, in order to reach a shady 
spot. In a word, they shewed no deficiency of vision whatever, and 
in all their behaviour exhibited the strongest contrast to frogs which 
had been really blinded by extirpation of the eye-ball or otherwise. 
It need hardly be added that care was taken to avoid any errors due 
to the frogs being guided by the heat as distinguished from the light 
of the sun. 

The idea naturally suggested itself that the frogs with bleached 
retinas, though not blind as regards the amount and intensity of 
light as a whole, might still be colour-blind. But an ingenious 
experiment proved this idea to be baseless. Kuhne finds that the 
vast majority of frogs (a few individual exceptions being from time 
to time met with) shew a most decided preference for green as com- 
pared with blue light. When a number of frogs are placed in a box, 
half of which is covered by green glass, and the other half with blue 
glass, they will in a very short time be all found huddled together 
under the green glass. Of course the experiment must be conducted 
in such a way that the illuminating power of the light and the tem- 
perature are nearly as possible the same in the two kinds of light. 
Erogs whose retinas have been bleached congregate under 
the green glass just in the same way as those whose 
retinas are in a normal condition. Very marked here also 
ia the contrast between frogs with bleached retinas and really blind 
frogs. Of course in all these experiments care was taken, by repeated 
examination of specimens, to make sure that no regeneration of 
purple took place during the experiments. 

The foregoing experiments prove at all events that, in frogs at 

* " Sehen ohne Sehpiu-pm-. " Untersuch. Bd. i. Hft. 



-*^PP] VISUAL PUEPLE. 103 

least, the presence of a store of visual purple in the rods is not 
essential to the appreciation of differences in colour. They are, it 
may be remarked, quite consistent with the favourite hypothesis 
that the cones only, and not the rods, are concerned in colour vision ; 
and they are at least highly suggestive of future researches. It would 
for instance be very interesting to ascertain the behaviour, when 
their retinas are bleached, of those animals, if there be any, which 
possess no cones, or rather in which all the terminal organs of the 
retina are coloured with the purple. 

Note F. p. 93. 
The Retinal Epithelium. 

The necessity in their various experiments of obtaining the retina 
free from the pigment epithelium has led Kuhne and Ewaldt 
to pay particular attention to the behaviour of the latter structure 
under the influence of light and other circumstances. These results 
are briefly as follows. 

In the eyes of frogs kept in the dark the pigment is for the most 
part concentrated in the body of the cell, a small quantity only 
being seated between the rods and reaching down for not more 
than about one-third the length of the outer limbs. After the 
animal has been exposed to Hght, the pigment is found, to a very 
large extent, arranged in strata between the rods. There is a 
tolerably large deposit between the posterior ends of the outer 
limbs; then follows a gap tolerably free from pigment, in front of 
which, between the anterior ends of the outer limbs, there is again 
a large deposit. From this it would appear that light acts as 
a stimulus to the cells of the retinal epithelium very much in the 
same way as it does to the cutaneous pigment-cells of the chameleon : 
they are by it excited to throw out pseudopodic processes which 
pass dowTi between the outer limbs of the rods. Upon the re- 
moval of the light, the processes are withdrawn towards or into the 
body of cell. Under red light, singularly enough, the forward 
transit of the pigment is even greater than in white light, a very 
large amount indeed of the pigment becoming heaped up at the 
front ends of the rods, the epithelium in consequence adhering very 
pertinaciously to the retina, so that the two are separated with the 
greatest difficulty. 

In general, exposure to light increases, and the withdrawal of 
light diminishes, the adhesion of the epithelium to the retina; but 
t Untersuch. Bd. i. Hft. 4, p, 411. 



104 VISUAL PURPLE. [APP. 

the absence or presence of light is not the only conditioning factor. 
Urari, apparently by giving rise to an oedema, favours the separation 
of the epithelium, while a low temperature produces the same effect 
as light, a high temperature having of course the oppo-^ite effect. A 
high temperature (37°0.), combined with urari poisoning, is sufficiently 
efficacious to overcome even the action of light; and sucli a treatment 
is found very useful in -experiments where it is desired to obtain 
retinas free from epithelium. 

Note G. p, 36. 
Action of Light on the Structure of the Rods. 

Kiihne and von Hornbostel* have satisfied themselves that 
in the frog light does produce an effect on the rods themselves. 
These swell vip under the influence of light during life, and thus, by 
"jamming up," so to speak, the processes of the retinal epithelium 
between, help to increase the cohesion of the retina to the epithelium. 

Note H. p. 46. 
Chemical Characters of Visual Purple. 

With regard to the chemical composition of visual purple we at 
present know little. Kuhnet has failed to discover in it any satis- 
factory evidence of iron being present in it, and it is not diffusable. 
It is not soluble in urea solutions or in melted paraffin ; and though 
easily destroyed by chlorine, nitrous acid, &c. it resists largely the 
action of ozone, permanganate of potash, ifec. 

When exposed, in the dark, to a temperature of 50° or 52° C, 
bleaching begins, becoming very rapid as the temperature rises to 
70°, taking place indeed almost instantaneously at 74° 0. The 
bleaching seems to be identical with that produced by light, in so far 
at least as the visible, changes are the same, the purple passing 
through chamois and yellow to white. 

Temperature has also a remarkable influence on the rapidity with 
which bleaching is effected by light, the purple becoming remark- 
ably sensitive at about 45°. 

It is possible that the change of the purple into yellow, and of 
the yellow into white, may be instances of dehydration. 

There are some facts which indicate that the visual purple, like 
hemoglobin, differs somewhat in nature in different animals. 

* Untersuch. Bd. i. Hft. 4. f Vntersuch. Bd. i. Hft. 4, p. 438. 

CAMBRIDGE : PEINTEB BY C. J. CLAY, M.A. AT THE UNIVERSITY PRESS.